diff --git a/CHANGELOG.md b/CHANGELOG.md
index 5ddaed3..85a8228 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -40,6 +40,26 @@ observable behavior and compatibility, not every internal refactor.
   (`qoder, codex, claude, cursor, qwen, copilot, pi`) when it rejects an
   unsupported `--platform`, matching the session-analysis and asset-baseline
   gates. The existing error prefix and exit behavior are unchanged.
+- Core Change Watch now requires framework-specific evidence before labeling
+  Rails or FastAPI, exposes bounded root Just recipes as statically discovered
+  unverified argv entrypoints, and keeps historical-only files out of current
+  recommended reads and action targets.
+- Evidence bundles now discover and privacy-filter one frozen Session population
+  before either Session facts or lead analysis hydrates it. Versioned redacted
+  bindings fail closed on population, selection, or admission contradictions
+  while preserving bounded lead selection and explicit zero-signal filtering.
+- Self-contained HTML reports now expose every fluency-dimension score track as
+  a labeled progressbar with a zero-to-100 range and the displayed rounded
+  score. Report validation rejects incomplete, duplicated, invalid, or
+  score-mismatched dimension progressbar contracts.
+- Chinese self-contained HTML reports now use standards-based language
+  segmentation to keep bounded word-like phrases together while preserving
+  normal wrapping around Latin text, paths, URLs, and longer content. Runtimes
+  without segmentation support fall back to readable escaped text, and English
+  reports remain unchanged.
+- HTML Evidence cards now display machine-owned Task Episode coverage from a
+  summary-facts companion, with legacy at-a-glance coverage retained only as a
+  compatibility fallback.
 
 ## 0.3.0 - 2026-07-27
 
diff --git a/docs/specs/2026-07-30-core-change-watch-profile-and-current-paths.md b/docs/specs/2026-07-30-core-change-watch-profile-and-current-paths.md
new file mode 100644
index 0000000..40fc9a2
--- /dev/null
+++ b/docs/specs/2026-07-30-core-change-watch-profile-and-current-paths.md
@@ -0,0 +1,95 @@
+# Core Change Watch Profile And Current Paths
+
+## Traceability
+
+- Spec ID: 2026-07-30-core-change-watch-profile-and-current-paths
+- Status: Implemented
+- Issue: None; focused evidence-correctness maintenance discovered by a validated Better Harness report.
+
+## Intent
+
+Keep Core Change Watch's static project profile and current-work guidance
+truthful. Framework labels must require framework-specific evidence, supported
+root Just recipes must be visible without being executed, and current reads or
+actions must not point to files that exist only in bounded Git history.
+
+## Acceptance Scenarios
+
+- AC-1: A Python project with FastAPI declared in a supported Python manifest
+  and generic `app/api` or `app/services` paths is identified as FastAPI, is not
+  identified as Rails, and receives no TypeScript-only path reason. A real
+  Rails fixture remains identified from Rails-specific evidence.
+- AC-2: Public recipes in a root `justfile`, `Justfile`, or `.justfile` are
+  represented additively as entry candidates with `kind: "just-recipe"`, an
+  argv-style `command`, the source path, and `executionStatus: "unverified"`.
+  Parsing is static, bounded, handles LF and CRLF, ignores private recipes, and
+  never runs Just or a shell.
+- AC-3: `recommendedReads` and every `followUpActions[].files` entry resolve to
+  a currently present tracked file or a currently present changed/untracked
+  file. Deleted historical paths remain available in raw history evidence but
+  are not projected as current work targets.
+- AC-4: Focused Core Change Watch tests, repository documentation checks, and
+  the full package gate pass without a new dependency.
+
+## Non-Goals
+
+- Session lane/lead population binding or source-fingerprint changes.
+- Renderer progress-track ARIA changes.
+- Agent Work Loop score changes or synthetic Task Episodes.
+- Executing Just recipes, shell-string dispatch, or proving command success.
+- Adding dependencies or a general TOML/Just parser.
+- Editing generated reports, installed plugin caches, Session content, Memory,
+  or user-home data.
+- Rewriting or deleting valid raw Git-history evidence.
+
+## Plan And Tasks
+
+Allowed files, copied verbatim from the approved Worker package:
+
+1. `docs/specs/2026-07-30-core-change-watch-profile-and-current-paths.md`
+2. `scripts/core-change-watch/project-profile.mjs`
+3. `scripts/core-change-watch/core-candidates.mjs`
+4. `scripts/core-change-watch/evidence-pack.mjs`
+5. `test/core-change-watch.test.mjs`
+6. `references/project-harness/core-change-watch.md`
+7. `CHANGELOG.md`
+8. `docs/better-harness-doc-links.mmd`
+
+Tasks:
+
+1. Add focused regression fixtures for AC-1 through AC-3 and preserve the
+   defect-specific failing run before production edits.
+2. Tighten framework/path evidence and add bounded static Just recipe
+   projection without changing existing entry-candidate meanings.
+3. Filter history-derived current-work projections through present repository
+   inventory while preserving the history profile.
+4. Update the canonical Core Change Watch guidance and changelog, regenerate
+   the Markdown routing graph, and mark this spec Implemented only after all
+   acceptance evidence passes.
+5. Run the change-traceability Review Readiness Check before commit or review.
+
+## Test And Review Evidence
+
+- The focused red run failed all three new regressions for the intended causes:
+  missing FastAPI evidence, missing Just recipe projection, and a deleted
+  historical path projected as current work. The same run passed 3/3 after the
+  implementation.
+- The real `database-caching` projection reports only `fastapi`, exposes
+  `just sync`, `just api-dev`, `just health`, and `just check` with unverified
+  status, uses no TypeScript reason for `app/api`, and emits only readable
+  recommended/action paths.
+- AC-1 to AC-3 red/green:
+  `node --test --test-name-pattern='framework-specific evidence|Just recipes|non-current historical paths' test/core-change-watch.test.mjs`
+- Focused suite: `node --test test/core-change-watch.test.mjs`
+- Real target projection: run `analyzeProjectProfile`,
+  `analyzeCoreCandidates`, and `buildEvidencePack` against `database-caching`;
+  assert FastAPI/no Rails, argv Just entries with unverified status, neutral
+  `app/api` reasons, and readable current-work paths.
+- Documentation: regenerate the routing graph and run
+  `node --test test/doc-link-graph.test.mjs`.
+- Package gate: `TMPDIR=<clean-temp-root> npm run check` passed with 884/884
+  tests plus npm-package and runtime-zip verification. The clean temp root
+  avoids unrelated `/tmp/CLAUDE.md` ancestor-instruction contamination.
+- Risk review: verify additive output fields, cross-platform path handling,
+  preserved Rails and raw-history fixtures, no shell execution, no dependency
+  changes, and an explicit AI co-author marker at commit time.
diff --git a/docs/specs/2026-07-30-html-cjk-line-breaking.md b/docs/specs/2026-07-30-html-cjk-line-breaking.md
new file mode 100644
index 0000000..5586285
--- /dev/null
+++ b/docs/specs/2026-07-30-html-cjk-line-breaking.md
@@ -0,0 +1,95 @@
+# Keep Chinese phrases together in HTML reports
+
+## Traceability
+
+- Spec ID: 2026-07-30-html-cjk-line-breaking
+- Status: Implemented
+- Issue: None; focused follow-up to the reviewed HTML progressbar visual QA.
+
+## Intent
+
+Keep ordinary Chinese word-like phrases together when the self-contained HTML
+report wraps headings and card copy. Segment visible `zh-CN` text with the
+standards-based `Intl.Segmenter` API and emit bounded, nonvisual keep-together
+spans so the result is deterministic across supported browsers. The generic
+renderer must improve reviewed Chinese content without hardcoded phrases,
+report-data mutation, runtime measurement, or a visual redesign.
+
+## Acceptance Scenarios
+
+- AC-1: A `zh-CN` HTML report uses `Intl.Segmenter` word segmentation for
+  visible text and wraps each ordinary all-Han word-like segment of at most
+  eight code points in `<span class="cjk-phrase">`. Representative phrases
+  including `口径`, `入口`, `结论`, and `命令` receive that deterministic markup.
+- AC-2: Phrase markup remains scoped to visible Chinese report text. English
+  reports retain their current markup, wrapping behavior, and visible copy;
+  attributes, accessibility labels, copied prompts, and embedded report JSON
+  remain plain escaped text without phrase spans or invisible joiners.
+- AC-3: Chinese headings, card titles, summaries, reasons, and ordinary
+  paragraphs continue to wrap between marked segments. Each keep-together span
+  is bounded to eight Han code points. Long tokens, paths, URLs, mixed Latin
+  text, and unrecognized segments remain escaped readable text and cannot gain
+  an unbounded no-wrap container.
+- AC-4: The existing dimension progressbar role, label, range, rounded value,
+  count validation, CSS, score computation, report data, and artifact set remain
+  unchanged.
+- AC-5: Focused HTML tests, the full package check, whitespace validation, and
+  the allowed-file audit pass. Before review completion, an environment with a
+  browser runtime inspects the same report at all three target widths.
+
+## Compatibility Boundary
+
+`Intl.Segmenter` is used only while generating the self-contained report. The
+renderer accepts only `isWordLike` segments that consist entirely of Han code
+points and are between two and eight code points long. All other segments use
+the existing HTML escaping path. If `Intl.Segmenter` is absent or throws, the
+entire value falls back to the same escaped readable text. CSS applies
+`white-space: nowrap` only to the bounded phrase spans, leaving normal wrapping
+available between spans and around Latin tokens, punctuation, paths, and URLs.
+
+## Non-goals
+
+- Hardcoding known report phrases or inserting word-joiner characters into
+  report data, visible copy, accessibility text, or copied prompts.
+- Preventing ordinary Chinese paragraphs from wrapping or forcing a complete
+  sentence, heading, or card onto one line.
+- Adding overflow, clipping, ellipsis, hidden content, fixed widths, custom
+  fonts, browser-specific JavaScript, dependencies, or generated artifacts.
+- Changing English layout, scores, findings, dimensions, report schemas,
+  progressbar semantics, interactions, or artifact names.
+
+## Plan and Tasks
+
+1. Add a focused renderer regression for deterministic markup, escaping,
+   bounded long-token handling, mixed Latin/path/URL preservation, English
+   preservation, unavailable-segmenter fallback, and the four reviewed phrases.
+   Record the missing-markup failure before renderer implementation. (AC-1..AC-4)
+2. Add one visible-text renderer using `Intl.Segmenter` and a bounded
+   `.cjk-phrase` keep-together rule. Route visible report copy through it while
+   leaving attributes, copied prompts, and serialized data unchanged. Remove
+   the insufficient browser-dependent `auto-phrase` contract. (AC-1..AC-4)
+3. Update the public changelog and verify the exact reviewed Chinese report at
+   375, 768, and 1280 pixels, including wrapping, overflow, clipping, console,
+   and preserved progressbar semantics. (AC-3..AC-5)
+
+## Test and Review Evidence
+
+- The initial focused test failed for the intended reason: the reviewed Chinese
+  phrases had no deterministic keep-together markup. It passed after the
+  standards-based segment renderer and bounded phrase span were added.
+- Browser QA then exposed descendant-selector leakage: nested phrase spans in
+  the score orbit, metrics, and evidence grid inherited block, width, and label
+  typography rules. A focused regression failed on the missing direct-child
+  selector contract before those existing rules were scoped to their direct
+  label and value children; nested phrase spans now remain inline.
+- The combined CJK and preserved progressbar suite passed 2/2:
+  `node --test --test-name-pattern='HTML CJK phrase|HTML dimension progressbar' test/harness-report-render-cli.test.mjs`.
+- The target-owned HTML suite passed 6/6:
+  `node --test --test-name-pattern='HTML' test/harness-report-render-cli.test.mjs`.
+- The coordinating PM ran the complete test suite serially with an isolated
+  temporary root: all 894 tests passed, followed by successful npm and runtime
+  zip package verification. This resolved the parallel temporary-ancestor
+  interference observed in two earlier Worker runs.
+- The exact reviewed Chinese report retains its source scores, finding count,
+  progressbar count, embedded report data, and artifact set. Post-fix browser
+  review at 375, 768, and 1280 pixels remains the final visual gate.
diff --git a/docs/specs/2026-07-30-html-dimension-progressbar-semantics.md b/docs/specs/2026-07-30-html-dimension-progressbar-semantics.md
new file mode 100644
index 0000000..8f0bbfd
--- /dev/null
+++ b/docs/specs/2026-07-30-html-dimension-progressbar-semantics.md
@@ -0,0 +1,76 @@
+# Give HTML dimension tracks progressbar semantics
+
+## Traceability
+
+- Spec ID: 2026-07-30-html-dimension-progressbar-semantics
+- Status: Implemented
+- Issue: None; focused accessibility-contract maintenance from a reviewed Better Harness finding.
+
+## Intent
+
+Make every fluency-dimension score track in the self-contained HTML report
+expose the score that sighted readers already see to assistive technology. The
+HTML validator must keep that semantic projection bound to the reviewed
+dimension count and rounded score so incomplete or stale markup cannot pass
+report validation.
+
+## Acceptance Scenarios
+
+- AC-1: Every rendered fluency-dimension card contains exactly one track with
+  `role="progressbar"`, a non-empty accessible label derived from the displayed
+  dimension label, `aria-valuemin="0"`, `aria-valuemax="100"`, and an
+  `aria-valuenow` equal to the integer score displayed beside that track.
+- AC-2: The HTML report validator passes the canonical rendered dimension
+  tracks and rejects a missing or extra dimension progressbar, a progressbar
+  with a missing or invalid role, label, minimum, maximum, or current value,
+  and a current value that differs from the reviewed rounded score.
+- AC-3: The renderer preserves existing CSS, layout, visible copy, dimension
+  ordering, score clamping and rounding, visual fill width, artifact names,
+  and generated report ownership.
+- AC-4: The focused regression, full renderer CLI suite, documentation graph,
+  doc-link suite, full package check, whitespace check, and allowed-file audit
+  pass from the assigned repository state.
+
+## Non-goals
+
+- Redesigning the dimension cards, tracks, responsive layout, color, or CSS.
+- Changing visible labels, summaries, localization, score computation,
+  clamping, rounding, or report data schemas.
+- Editing generated reports, templates, installed plugin caches, dependencies,
+  custom SVG, or any report mode other than the canonical HTML renderer.
+- Adding browser interaction, live-region announcements, or a new artifact.
+
+## Plan and Tasks
+
+1. Add one focused regression test that asserts the rendered per-dimension
+   progressbar contract and mutation-based validator rejection for missing,
+   invalid, count-mismatched, and score-mismatched semantics. Record the test
+   failing against the current renderer before implementation. (AC-1, AC-2)
+2. Reuse the renderer's existing display label and rounded score when emitting
+   the complete progressbar attributes, without changing track CSS or fill
+   width. (AC-1, AC-3)
+3. Extend `evaluateHtmlReport` to bind exactly one valid progressbar to each
+   reviewed dimension card and reject semantic count or score drift. (AC-2)
+4. Update the public changelog, regenerate the Markdown routing graph required
+   for the new spec, run every assigned verification command, and mark this
+   spec Implemented only after the evidence passes. (AC-4)
+
+## Test and Review Evidence
+
+- The required red run failed for the intended contract reason: the current
+  renderer produced zero semantic progressbars for five reviewed dimensions
+  (`0 !== 5`). The same focused command passed after implementation:
+  `node --test --test-name-pattern='dimension progressbar' test/harness-report-render-cli.test.mjs`.
+- The full renderer suite passed 15/15:
+  `node --test test/harness-report-render-cli.test.mjs`.
+- Documentation graph generation and all six doc-link checks passed:
+  `node scripts/doc-link-graph/cli.mjs skills/better-harness && node --test test/doc-link-graph.test.mjs`.
+- The complete package gate passed from a clean temporary root under
+  `/var/tmp` with 893 tests and successful npm/runtime-zip package verification:
+  `TMPDIR=/var/tmp/better-harness-t016.lxCc80 npm run check`.
+- Scope and whitespace:
+  `git diff --check -- <allowed files>` plus an allowed-file-only status audit.
+- Risk review: the focused test removes and duplicates a progressbar, mutates
+  each semantic attribute class, and changes `aria-valuenow`; all mutations
+  fail validation while canonical output passes. The CSS and existing visual
+  fill-width expression are unchanged.
diff --git a/docs/specs/2026-07-30-html-summaryfacts-episode-coverage.md b/docs/specs/2026-07-30-html-summaryfacts-episode-coverage.md
new file mode 100644
index 0000000..d35b820
--- /dev/null
+++ b/docs/specs/2026-07-30-html-summaryfacts-episode-coverage.md
@@ -0,0 +1,60 @@
+# Preserve summary-facts Episode coverage in HTML
+
+## Traceability
+
+- Spec ID: 2026-07-30-html-summaryfacts-episode-coverage
+- Status: Implemented
+- Issue: None; focused renderer correctness repair found during final browser QA.
+
+## Intent
+
+Make the self-contained HTML Evidence section display the machine-owned Episode
+coverage already embedded in a split findings plus summary-facts companion
+render. The current renderer preserves those facts in report data but reads only
+the legacy `atAGlance.coverage` projection for its visible cards, producing a
+false zero when the canonical `evidenceBoundary.episodeCoverage` is present.
+
+## Acceptance Scenarios
+
+- AC-1: When `summary.evidenceBoundary.episodeCoverage` reports 14 total and 12
+  edited Episodes, the visible HTML Evidence cards display 14 and 12.
+- AC-2: Machine-owned `evidenceBoundary.episodeCoverage` takes precedence over
+  stale or absent legacy `atAGlance.coverage` values.
+- AC-3: A legacy report that has only `atAGlance.coverage` keeps its existing
+  visible Episode counts.
+- AC-4: No score, finding, schema, CSS, layout, CJK, interaction, or artifact
+  ownership behavior changes.
+- AC-5: Focused renderer tests, the full HTML suite, the complete serial package
+  suite, package verification, whitespace checks, and final browser QA pass.
+
+## Non-goals
+
+- Changing evidence collection, Session population binding, Episode admission,
+  report scores, findings, or summary-facts schemas.
+- Adding fallback inference from usage counts or visible prose.
+- Editing generated reports or installed plugin caches.
+- Redesigning the Evidence section or adding new UI.
+
+## Plan and Tasks
+
+1. Add a focused renderer regression for canonical summary-facts precedence and
+   legacy compatibility, and preserve its red result. (AC-1, AC-2, AC-3)
+2. Read Episode coverage from the canonical machine-facts owner with the legacy
+   projection as a compatibility fallback. (AC-1, AC-2, AC-3, AC-4)
+3. Run the assigned verification and record the evidence before marking this
+   spec Implemented. (AC-5)
+
+## Test and Review Evidence
+
+- The focused red run failed because the canonical 14/12 summary-facts values
+  rendered as the conflicting legacy 1/1 values. The unchanged focused command
+  passed after implementation:
+  `node --test --test-name-pattern='HTML evidence episode coverage' test/harness-report-render-cli.test.mjs`.
+- The complete HTML-focused renderer suite passed 7/7:
+  `node --test --test-name-pattern='HTML' test/harness-report-render-cli.test.mjs`.
+- The isolated serial package suite passed 895/895:
+  `TMPDIR=/var/tmp/better-harness-t025.dES6tD npm test -- --test-concurrency=1`.
+- Package verification passed with 324 npm entries and 348 runtime-zip entries:
+  `TMPDIR=/var/tmp/better-harness-t025.dES6tD npm run pack:verify`.
+- Final report rendering, three-width browser QA, scope, whitespace, and review
+  evidence are recorded in the contribution receipt before commit.
diff --git a/docs/specs/2026-07-30-session-population-binding.md b/docs/specs/2026-07-30-session-population-binding.md
new file mode 100644
index 0000000..18d7fe5
--- /dev/null
+++ b/docs/specs/2026-07-30-session-population-binding.md
@@ -0,0 +1,119 @@
+# Session Population Binding
+
+## Traceability
+
+- Spec ID: 2026-07-30-session-population-binding
+- Status: Implemented
+- Issue: None; focused evidence-integrity maintenance discovered by a validated Better Harness report.
+
+## Intent
+
+Make one evidence-bundle run derive Session facts and lead analysis from one
+frozen, privacy-filtered Session population. The binding must prevent an active
+Session from entering the lead after it was omitted from facts, without
+persisting or exposing the private inventory or weakening bounded lead
+selection.
+
+## Acceptance Scenarios
+
+- AC-1: The bundle discovers the frozen Session population once, removes Qoder
+  home-only sources and an exact explicit/provider active Session identity
+  before either lane hydrates it, and supplies the same eligible snapshot to
+  both collectors. With an explicit frozen `until`, exact identity exclusion
+  remains enabled while recency-window inference remains disabled.
+- AC-2: Public output contains only bounded, versioned bindings: frozen scope
+  and omission policy/counts, eligible population count/fingerprint, Session
+  all-eligible selection count/fingerprint, and lead bounded selection
+  count/fingerprint with a parent-population assertion. It contains no raw
+  Session IDs, source paths, prompts, commands, content, Memories, user-home
+  payloads, reversible locators, or serialized snapshot.
+- AC-3: Bundle validation fails closed with stable redacted error codes when a
+  binding is missing or mismatched, a selected fingerprint is not a subset of
+  its population, or admission counters do not reconcile. Session admission
+  satisfies `taskEpisodes = candidateEpisodes + noRequest + selfAnalysis +
+  lowSignal`, `candidateEpisodes = distinctRequests + duplicateRequests`, and
+  `distinctRequests = emittedCandidates + candidateBudget`.
+- AC-4: Lead admission satisfies `projectedEpisodes = admittedEpisodes +
+  zeroSignalDiscardedEpisodes` and `admittedEpisodes = retained taskEpisodes`.
+  Selected fingerprints remain lane-specific for bounded strategies; cross-lane
+  Episode totals are compared only when selected and projection-policy
+  fingerprints match. Legitimate zero-signal filtering remains valid.
+- AC-5: Existing scoring, lead stratification and limits, raw-data exclusions,
+  provider adapters, and standalone analysis commands remain unchanged.
+
+## Non-Goals
+
+- Reading or migrating raw Sessions, Memories, user-home content, private
+  plugin data, or unsanitized generated reports.
+- Persisting the shared population snapshot or exposing private identity data.
+- All-eligible lead hydration, larger limits, weaker stratification, synthetic
+  Task Episodes, or score/weight changes.
+- Renderer accessibility work, Core Change Watch work, report edits, installed
+  plugin-cache changes, releases, merges, or production publication.
+- Provider-adapter changes or a new dependency.
+
+## Plan And Tasks
+
+Allowed files, copied verbatim from the approved Worker package:
+
+1. `docs/specs/2026-07-30-session-population-binding.md`
+2. `scripts/session-analysis/session-population.mjs`
+3. `scripts/harness-analysis/evidence-bundle/contract.mjs`
+4. `scripts/harness-analysis/evidence-bundle/index.mjs`
+5. `scripts/harness-analysis/evidence-bundle/session-evidence.mjs`
+6. `scripts/harness-analysis/report-run.mjs`
+7. `scripts/harness-analysis/task-loop-source.mjs`
+8. `test/session-population.test.mjs`
+9. `test/better-harness-evidence-bundle.test.mjs`
+10. `test/task-loop-source.test.mjs`
+11. `CHANGELOG.md`
+12. `docs/better-harness-doc-links.mmd`
+
+Tasks:
+
+1. Add sanitized fixtures for exact active-session exclusion, shared snapshot
+   reuse, binding mismatch, admission reconciliation, and zero-signal filtering.
+2. Preserve a test-only red run that demonstrates active-session divergence and
+   missing fail-closed binding before production edits.
+3. Implement the private population owner, public binding contract, shared
+   orchestration, and bounded lead projection without persistence or new data
+   authority.
+4. Regenerate the Markdown routing graph, update the changelog, run focused and
+   full gates, and mark this spec Implemented only after evidence passes.
+5. Perform Review Readiness over the no-issue rationale, Spec, Tests, Risk, AI,
+   Refs, generated files, and staged/unstaged boundaries before commit.
+
+## Test And Review Evidence
+
+- Red run: six focused tests failed for the intended missing owner, unshared
+  population, non-failing mismatch, and unavailable zero-signal accounting
+  before production edits.
+- Focused binding tests: nine population, binding, active-session, count
+  contradiction, frozen-inventory, and zero-signal tests pass.
+- Full Session/bundle/source compatibility: 61/61 tests pass across
+  `test/session-population.test.mjs`, `test/better-harness-evidence-bundle.test.mjs`,
+  `test/task-loop-source.test.mjs`, and `test/harness-report-run.test.mjs`.
+- Real target: a normal-depth Codex bundle for `database-caching` completed with
+  both lanes available and the binding `bound`. It recorded exact active
+  identity availability, omitted one active Session, bound four eligible
+  Sessions to the same parent fingerprint, and reconciled four projected
+  Episodes as four explicit zero-signal discards and zero retained Episodes.
+- Privacy serialization audit: the focused bundle fixture proves neither the
+  private inventory nor its Session IDs appear in serialized output; the real
+  target inspection printed only versioned bindings and aggregate counters.
+- Review hardening: the lead source path now skips a second `sources` discovery
+  when a frozen population is supplied, and the privacy regression checks
+  specific fixture identities instead of rejecting macOS `/private/tmp` paths.
+- Review validation: the focused Session/report suite passed 123/123 tests, the
+  repository gate passed 1006/1006 tests plus package verification, and browser
+  QA confirmed complete progressbar semantics with no console errors or page
+  overflow in the Chinese HTML fixture.
+- Documentation: the canonical routing graph regenerated without drift and the
+  six documentation checks pass.
+- Package gate: `TMPDIR=<clean-temp-root> npm run check` passed 892/892 tests
+  plus npm-package (321 entries) and runtime-zip (345 entries) verification.
+- Review Readiness: the diff is confined to the approved owner/test/spec/
+  changelog files; the optional contract and generated graph files were not
+  changed. No dependency, provider adapter, scoring owner, generated report,
+  installed cache, raw identity, or private payload changed. The commit must
+  carry exactly one required Codex co-author marker.
diff --git a/references/project-harness/core-change-watch.md b/references/project-harness/core-change-watch.md
index fb08aa0..3f52fbb 100644
--- a/references/project-harness/core-change-watch.md
+++ b/references/project-harness/core-change-watch.md
@@ -12,6 +12,25 @@ the resulting boundary evidence-qualified in visible output: say "inferred core
 boundary" or "affected core files", and name concrete source paths when they
 support the claim.
 
+## Project Profile And Current-Work Paths
+
+Framework labels require framework-specific manifest or convention evidence.
+Generic directories such as `app/api` and `app/services` remain useful
+language-neutral path signals, but they do not establish Rails, TypeScript, or
+another framework by themselves.
+
+A root `justfile`, `Justfile`, or `.justfile` contributes bounded public recipe
+entry candidates. These rows expose an argv-style command and retain
+`executionStatus: "unverified"`; collection parses the file statically and does
+not run Just or a shell. Private recipes remain outside the command surface.
+
+Raw bounded history may retain files that no longer exist because those paths
+remain valid historical facts. Before `recommendedReads` and follow-up action
+file lists are emitted, Core Change Watch intersects their candidates with the
+currently present tracked inventory plus currently present changed/untracked
+files. Do not interpret a historical hot path as a current edit target unless
+it passes that projection boundary.
+
 ## High-Impact Expansion
 
 Expand before synthesis for Core, platform, IDE plugin, LSP, auth, tool
diff --git a/scripts/core-change-watch/core-candidates.mjs b/scripts/core-change-watch/core-candidates.mjs
index 125dd29..3e4b8e9 100644
--- a/scripts/core-change-watch/core-candidates.mjs
+++ b/scripts/core-change-watch/core-candidates.mjs
@@ -40,7 +40,7 @@ const CORE_PATH_PATTERNS = [
   { pattern: /^cmd\/[^/]+/i, score: 14, reason: "go command entrypoint" },
   { pattern: /^src\/main\/java\//i, score: 24, reason: "java production source" },
   { pattern: /^packages\/[^/]+/i, score: 16, reason: "workspace package" },
-  { pattern: /^app\/api(\/|$)/i, score: 20, reason: "typescript api route" },
+  { pattern: /^app\/api(\/|$)/i, score: 20, reason: "api route path" },
   { pattern: /^src\/(controller|controllers|service|services|entity|repository|security)(\/|$)/i, score: 18, reason: "framework source path" },
   { pattern: /^app\/(http|models|services|providers|policies|jobs|console|controllers|models)(\/|$)/i, score: 18, reason: "framework app path" },
   { pattern: /^routes(\/|$)/i, score: 12, reason: "framework route path" },
@@ -48,6 +48,7 @@ const CORE_PATH_PATTERNS = [
 ];
 
 const FRAMEWORK_PATH_SIGNALS = [
+  { framework: "fastapi", pattern: /^app\/api(\/|$)/, score: 18, reason: "fastapi route path" },
   { framework: "laravel", pattern: /^app\/(Http|Models|Services|Providers|Policies|Jobs|Console)(\/|$)/, score: 22, reason: "laravel application path" },
   { framework: "laravel", pattern: /^routes(\/|$)/, score: 14, reason: "laravel route path" },
   { framework: "symfony", pattern: /^src\/(Controller|Entity|Service|Security|Repository|MessageHandler)(\/|$)/, score: 22, reason: "symfony application path" },
diff --git a/scripts/core-change-watch/evidence-pack.mjs b/scripts/core-change-watch/evidence-pack.mjs
index f8c56d4..e251298 100644
--- a/scripts/core-change-watch/evidence-pack.mjs
+++ b/scripts/core-change-watch/evidence-pack.mjs
@@ -1,5 +1,8 @@
 #!/usr/bin/env node
 
+import { existsSync } from "node:fs";
+import path from "node:path";
+
 import {
   DEFAULT_MAX_COMMITS,
   SCHEMA_VERSION,
@@ -8,6 +11,7 @@ import {
   isCli,
   isSupportingFile,
   languageFor,
+  listTrackedFiles,
   normalizeHistoryWindows,
   normalizeLanguages,
   option,
@@ -23,8 +27,8 @@ import { analyzeDiffImpact } from "./diff-impact.mjs";
 import { analyzeGitHistoryProfile } from "./git-history-profile.mjs";
 import { analyzeProjectProfile } from "./project-profile.mjs";
 
-function addRead(reads, path, reason, priority = 50) {
-  if (!path) {
+function addRead(reads, path, reason, priority = 50, currentPaths = null) {
+  if (!path || (currentPaths && !currentPaths.has(path))) {
     return;
   }
   const role = fileRoleFor(path);
@@ -45,32 +49,32 @@ function addRead(reads, path, reason, priority = 50) {
   });
 }
 
-function recommendedReads(core, diff, profile, maxReads) {
+function recommendedReads(core, diff, profile, maxReads, currentPaths) {
   const reads = new Map();
 
   for (const file of diff.changedFiles) {
-    addRead(reads, file.path, "changed file", 100);
+    addRead(reads, file.path, "changed file", 100, currentPaths);
   }
 
   for (const hit of diff.coreHits) {
-    addRead(reads, hit.filePath, `changed core candidate ${hit.path}`, 95);
+    addRead(reads, hit.filePath, `changed core candidate ${hit.path}`, 95, currentPaths);
   }
 
   for (const hit of diff.hotHits) {
-    addRead(reads, hit.path, "changed hot file", 90);
+    addRead(reads, hit.path, "changed hot file", 90, currentPaths);
   }
 
   for (const candidate of core.candidates.slice(0, 8)) {
     for (const file of candidate.evidence.hotFiles.slice(0, 3)) {
-      addRead(reads, file.path, `hot file under ${candidate.path}`, 75);
+      addRead(reads, file.path, `hot file under ${candidate.path}`, 75, currentPaths);
     }
     for (const file of candidate.evidence.sourceFiles.slice(0, 2)) {
-      addRead(reads, file, `representative source under ${candidate.path}`, 55);
+      addRead(reads, file, `representative source under ${candidate.path}`, 55, currentPaths);
     }
   }
 
   for (const entry of profile.entryCandidates.slice(0, 5)) {
-    addRead(reads, entry.path, "entry candidate", 45);
+    addRead(reads, entry.path, "entry candidate", 45, currentPaths);
   }
 
   return [...reads.values()]
@@ -78,12 +82,12 @@ function recommendedReads(core, diff, profile, maxReads) {
     .slice(0, maxReads);
 }
 
-function actionFiles(files, limit = 8) {
+function actionFiles(files, limit = 8, currentPaths = null) {
   const seen = new Set();
   const result = [];
   for (const file of files) {
     const path = typeof file === "string" ? file : file?.path;
-    if (!path || seen.has(path)) {
+    if (!path || seen.has(path) || (currentPaths && !currentPaths.has(path))) {
       continue;
     }
     seen.add(path);
@@ -99,29 +103,30 @@ function actionFiles(files, limit = 8) {
   return result;
 }
 
-function supportingFilesByRole(history, diff, role) {
+function supportingFilesByRole(history, diff, role, currentPaths) {
   const fromDiff = diff.companionHits?.filter((item) => item.role === role) ?? [];
   const fromHistory = history.supportingHotFiles?.filter((item) => item.role === role) ?? [];
-  return actionFiles([...fromDiff, ...fromHistory], 8);
+  return actionFiles([...fromDiff, ...fromHistory], 8, currentPaths);
 }
 
-function buildFollowUpActions(core, diff, history, changeDrift, maxActions) {
+function buildFollowUpActions(core, diff, history, changeDrift, maxActions, currentPaths) {
   const actions = [];
   const topCoreFiles = actionFiles(core.candidates.slice(0, 4).flatMap((candidate) => [
     ...candidate.evidence.hotFiles,
     ...candidate.evidence.sourceFiles,
-  ]), 10);
+  ]), 10, currentPaths);
   const changedPrimaryFiles = actionFiles(
     diff.changedFiles.filter((file) => !file.supporting),
     10,
+    currentPaths,
   );
-  const affectedCoreFiles = actionFiles(diff.coreHits.map((hit) => hit.filePath), 10);
-  const localizationFiles = supportingFilesByRole(history, diff, "localization");
-  const documentationFiles = supportingFilesByRole(history, diff, "documentation");
+  const affectedCoreFiles = actionFiles(diff.coreHits.map((hit) => hit.filePath), 10, currentPaths);
+  const localizationFiles = supportingFilesByRole(history, diff, "localization", currentPaths);
+  const documentationFiles = supportingFilesByRole(history, diff, "documentation", currentPaths);
   const testFiles = actionFiles([
     ...(history.supportingHotFiles?.filter((item) => item.role === "test" || item.role === "fixture") ?? []),
     ...diff.changedFiles.filter((item) => item.role === "test" || item.role === "fixture"),
-  ], 8);
+  ], 8, currentPaths);
 
   if (topCoreFiles.length > 0) {
     actions.push({
@@ -190,7 +195,7 @@ function buildFollowUpActions(core, diff, history, changeDrift, maxActions) {
       priority: finding.severity === "high" ? "high" : "medium",
       title: companionActionTitle(finding),
       reason: finding.evidence,
-      files: actionFiles([...finding.triggerFiles, ...finding.candidateCompanionFiles], 8),
+      files: actionFiles([...finding.triggerFiles, ...finding.candidateCompanionFiles], 8, currentPaths),
       passCheck: finding.action,
     });
   }
@@ -214,6 +219,14 @@ function buildFollowUpActions(core, diff, history, changeDrift, maxActions) {
   return actions.slice(0, maxActions);
 }
 
+function currentWorkPaths(repoRoot, diff) {
+  const candidates = new Set([
+    ...listTrackedFiles(repoRoot),
+    ...diff.changedFiles.map((file) => file.path),
+  ]);
+  return new Set([...candidates].filter((filePath) => existsSync(path.resolve(repoRoot, filePath))));
+}
+
 function stageFilters(projectProfile, historyProfile, coreAnalysis, diffImpact) {
   return {
     projectProfile: projectProfile.filters,
@@ -522,7 +535,15 @@ export async function buildEvidencePack(options = {}) {
     ignore: options.ignore,
     changedFiles: diffImpact.changedFiles,
   });
-  const followUpActions = buildFollowUpActions(coreAnalysis, diffImpact, historyProfile, changeDrift, maxFollowUpActions);
+  const currentPaths = currentWorkPaths(repoRoot, diffImpact);
+  const followUpActions = buildFollowUpActions(
+    coreAnalysis,
+    diffImpact,
+    historyProfile,
+    changeDrift,
+    maxFollowUpActions,
+    currentPaths,
+  );
   const filters = stageFilters(projectProfile, historyProfile, coreAnalysis, diffImpact);
   const pack = {
     schemaVersion: SCHEMA_VERSION,
@@ -567,7 +588,7 @@ export async function buildEvidencePack(options = {}) {
     coreAnalysis,
     diffImpact,
     changeDrift,
-    recommendedReads: recommendedReads(coreAnalysis, diffImpact, projectProfile, maxRecommendedReads),
+    recommendedReads: recommendedReads(coreAnalysis, diffImpact, projectProfile, maxRecommendedReads, currentPaths),
     followUpActions,
     agentGuidance: buildAgentGuidance(historyProfile, coreAnalysis, filters),
     reviewMatrix: buildReviewMatrix(projectProfile, historyProfile, coreAnalysis, diffImpact, changeDrift, followUpActions),
diff --git a/scripts/core-change-watch/project-profile.mjs b/scripts/core-change-watch/project-profile.mjs
index a80812c..fcb5114 100644
--- a/scripts/core-change-watch/project-profile.mjs
+++ b/scripts/core-change-watch/project-profile.mjs
@@ -57,6 +57,8 @@ const SOURCE_FILES_PER_AGENT_INSTRUCTION = 250;
 const SIGNIFICANT_SOURCE_ROOT_FILES = 100;
 const SOURCE_LINE_METHOD = "tracked readable primary source files only; excludes tests, non-source assets, generated/dependency paths, and untracked files";
 const SOURCE_LINE_SKIPPED_METHOD = "not measured by default; use --measure-source-lines to count tracked readable primary source files only";
+const ROOT_JUSTFILES = new Set(["justfile", "Justfile", ".justfile"]);
+const MAX_JUST_RECIPES = 100;
 
 function scriptNames(manifest, filePath) {
   if (!manifest?.scripts || typeof manifest.scripts !== "object") {
@@ -137,6 +139,41 @@ function entryScore(filePath) {
   return 0;
 }
 
+function justRecipeScore(name) {
+  const conventional = new Map([
+    ["default", 100],
+    ["sync", 99],
+    ["api-dev", 98],
+    ["health", 97],
+    ["check", 96],
+  ]);
+  return conventional.get(name) ?? 70;
+}
+
+function justRecipeNames(text) {
+  const recipes = [];
+  let privateRecipe = false;
+  for (const line of String(text ?? "").split(/\r?\n/u)) {
+    const trimmed = line.trim();
+    if (!trimmed || trimmed.startsWith("#")) continue;
+    if (!/^\s/u.test(line) && /^\[[^\]]+\]$/u.test(trimmed)) {
+      privateRecipe = /(?:^|[\s,])private(?:[\s,]|$)/u.test(trimmed.slice(1, -1));
+      continue;
+    }
+    const match = !/^\s/u.test(line)
+      ? /^@?([A-Za-z_][A-Za-z0-9_-]*)\b(?!\s*:=)[^:\r\n]*:(?:\s|$)/u.exec(line)
+      : null;
+    if (match) {
+      if (!privateRecipe && !match[1].startsWith("_")) recipes.push(match[1]);
+      privateRecipe = false;
+      if (recipes.length >= MAX_JUST_RECIPES) break;
+      continue;
+    }
+    if (!/^\s/u.test(line)) privateRecipe = false;
+  }
+  return [...new Set(recipes)];
+}
+
 function corePathHint(filePath, analysisScope) {
   const localDirectory = analysisDirectoryFor(toAnalysisRelativePath(filePath, analysisScope));
   if (/(^|\/)(core|auth|security|permission|permissions|crypto|session|payment|billing|domain|service|api)(\/|$)/i.test(localDirectory)) {
@@ -313,6 +350,7 @@ function detectFrameworks(repoRoot, trackedFiles) {
     || firstManifestWithDependency(composerManifests, "symfony/http-kernel");
   const codeIgniterManifest = firstManifestWithDependency(composerManifests, "codeigniter4/framework");
   const railsGemfile = firstTextMatch(repoRoot, fileSet, /(^|\/)Gemfile$/, /\bgem\s+["']rails["']/);
+  const railsRoutes = manifestFiles(fileSet, /(^|\/)config\/routes\.rb$/)[0] ?? "";
   const pythonManifestText = joinedManifestText(repoRoot, fileSet, /(^|\/)(requirements\.txt|pyproject\.toml|Pipfile)$/);
   const javaBuildText = joinedManifestText(repoRoot, fileSet, /(^|\/)(pom\.xml|build\.gradle|build\.gradle\.kts)$/);
 
@@ -331,8 +369,11 @@ function detectFrameworks(repoRoot, trackedFiles) {
   if (codeIgniterManifest || fileSet.has("app/Config/Routes.php")) {
     addFrameworkSignal(signals, "codeigniter", codeIgniterManifest ? 70 : 30, codeIgniterManifest ? `${codeIgniterManifest}:codeigniter4/framework` : "app/Config/Routes.php");
   }
-  if (railsGemfile || fileSet.has("config/routes.rb") || hasFileMatching(fileSet, /^app\/(models|controllers|services)\//)) {
-    addFrameworkSignal(signals, "rails", railsGemfile ? 75 : 35, railsGemfile ? `${railsGemfile}:rails` : "rails app paths");
+  if (railsGemfile || railsRoutes) {
+    addFrameworkSignal(signals, "rails", railsGemfile ? 75 : 45, railsGemfile ? `${railsGemfile}:rails` : railsRoutes);
+  }
+  if (/\bfastapi\b/iu.test(pythonManifestText)) {
+    addFrameworkSignal(signals, "fastapi", 75, "python manifest:fastapi");
   }
   if (/\bdjango\b/i.test(pythonManifestText) || fileSet.has("manage.py") || hasFileMatching(fileSet, /(^|\/)settings\.py$/)) {
     addFrameworkSignal(signals, "django", /\bdjango\b/i.test(pythonManifestText) ? 70 : 35, /\bdjango\b/i.test(pythonManifestText) ? "python manifest:django" : "manage.py/settings.py");
@@ -526,6 +567,10 @@ function buildProjectInfo({
       path: item.path,
       language: item.language,
       score: item.score,
+      ...(item.kind ? { kind: item.kind } : {}),
+      ...(item.command ? { command: item.command } : {}),
+      ...(item.sourcePath ? { sourcePath: item.sourcePath } : {}),
+      ...(item.executionStatus ? { executionStatus: item.executionStatus } : {}),
     })),
     manifests: manifests.slice(0, 8),
   };
@@ -565,6 +610,23 @@ export async function analyzeProjectProfile(options = {}) {
       continue;
     }
 
+    if (!normalized.includes("/") && ROOT_JUSTFILES.has(normalized)) {
+      const recipes = justRecipeNames(safeReadText(repoRoot, normalized));
+      manifests.push({ path: normalized, kind: "just", scripts: [...recipes].sort() });
+      for (const recipe of recipes) {
+        entryCandidates.push({
+          path: normalized,
+          language: "just",
+          score: justRecipeScore(recipe),
+          reasons: compactReasonList(["statically discovered root Just recipe"]),
+          kind: "just-recipe",
+          command: ["just", recipe],
+          sourcePath: normalized,
+          executionStatus: "unverified",
+        });
+      }
+    }
+
     if (isManifestFile(normalized)) {
       const manifest = readJsonFile(repoRoot, normalized);
       manifests.push({
diff --git a/scripts/harness-analysis/evidence-bundle/index.mjs b/scripts/harness-analysis/evidence-bundle/index.mjs
index 208f46f..2390dcb 100644
--- a/scripts/harness-analysis/evidence-bundle/index.mjs
+++ b/scripts/harness-analysis/evidence-bundle/index.mjs
@@ -1,4 +1,5 @@
 import { analyzeHarnessEvidence } from "../report-run.mjs";
+import { validateSessionPopulationBundle } from "../../session-analysis/session-population.mjs";
 import { resolveWorkspaceTopology } from "../../workspace-topology/index.mjs";
 import { collectAgentCustomize } from "./agent-customize.mjs";
 import {
@@ -11,7 +12,10 @@ import {
   unavailableLane,
 } from "./contract.mjs";
 import { collectProjectHarness } from "./project-harness.mjs";
-import { collectSessionEvidence } from "./session-evidence.mjs";
+import {
+  collectSessionEvidence,
+  collectSessionPopulation,
+} from "./session-evidence.mjs";
 
 async function capture(owner, operation) {
   try {
@@ -27,7 +31,7 @@ async function capture(owner, operation) {
   }
 }
 
-async function collectLead(context, options, analyze) {
+async function collectLead(context, options, analyze, sessionPopulation) {
   try {
     const data = await analyze({
       workspace: context.workspace,
@@ -36,6 +40,7 @@ async function collectLead(context, options, analyze) {
       since: context.window.since,
       until: context.window.until,
       format: "json",
+      sessionPopulation,
       "include-global-capabilities": context.authority.includeUserHome,
       ...(options["canvas-out"] ? { "canvas-out": options["canvas-out"] } : {}),
       ...(options["replace-canvas"] ? { "replace-canvas": options["replace-canvas"] } : {}),
@@ -54,6 +59,54 @@ async function collectLead(context, options, analyze) {
   }
 }
 
+function populationDiagnostics(population, sessionEvidence, lead) {
+  if (!population?.binding) {
+    return { status: "unavailable" };
+  }
+  const session = sessionEvidence?.data
+    ? {
+        population: sessionEvidence.data.populationBinding,
+        selection: sessionEvidence.data.selectionBinding,
+        admission: sessionEvidence.data.admissionBinding,
+      }
+    : null;
+  const leadBinding = lead?.data?.sessionBinding ?? null;
+  const errors = validateSessionPopulationBundle({
+    population: population.binding,
+    session,
+    lead: leadBinding,
+  });
+  const sessionEligibleCount = Number(sessionEvidence?.data?.scope?.eligibleSessions ?? -1);
+  const sessionSelectedCount = Number(sessionEvidence?.data?.scope?.selectedSessions ?? -1);
+  const leadSelection = lead?.data?.summaryFacts?.evidenceBoundary?.manifest?.selection ?? {};
+  if (sessionEligibleCount !== population.binding.eligible.count
+    || sessionSelectedCount !== session?.selection?.selected?.count) {
+    errors.push("Session public counts do not match its population binding");
+  }
+  if (Number(leadSelection.eligibleCount ?? -1) !== population.binding.eligible.count
+    || Number(leadSelection.analyzedCount ?? -1) !== leadBinding?.selection?.selected?.count) {
+    errors.push("lead public counts do not match its population binding");
+  }
+  const sessionAdmission = session?.admission ?? {};
+  const leadAdmission = leadBinding?.admission ?? {};
+  const comparable = session?.selection?.selected?.fingerprint === leadBinding?.selection?.selected?.fingerprint
+    && session?.selection?.projectionPolicyFingerprint === leadBinding?.selection?.projectionPolicyFingerprint;
+  return {
+    status: errors.length > 0 ? "conflict" : "bound",
+    population: population.binding,
+    sessionSelection: session?.selection ?? null,
+    leadSelection: leadBinding?.selection ?? null,
+    episodes: {
+      comparison: comparable ? "comparable" : "not-comparable-selection-or-policy",
+      sessionTaskEpisodes: Number(sessionAdmission.taskEpisodes ?? 0),
+      leadProjectedEpisodes: Number(leadAdmission.projectedEpisodes ?? 0),
+      leadRetainedEpisodes: Number(leadAdmission.retainedTaskEpisodes ?? 0),
+      leadZeroSignalDiscardedEpisodes: Number(leadAdmission.zeroSignalDiscardedEpisodes ?? 0),
+    },
+    ...(errors.length > 0 ? { errorCodes: ["SESSION_POPULATION_BINDING_MISMATCH"] } : {}),
+  };
+}
+
 export async function collectEvidenceBundle(options = {}, dependencies = {}) {
   const now = dependencies.now?.() ?? new Date();
   const baseContext = freezeEvidenceBundleContext(options, now);
@@ -73,12 +126,34 @@ export async function collectEvidenceBundle(options = {}, dependencies = {}) {
   const projectCollector = dependencies.collectProjectHarness ?? collectProjectHarness;
   const customizeCollector = dependencies.collectAgentCustomize ?? collectAgentCustomize;
   const leadAnalyzer = dependencies.analyzeHarnessEvidence ?? analyzeHarnessEvidence;
-  const [sessionEvidence, projectHarness, agentCustomize, lead] = await Promise.all([
-    capture("session-evidence", () => sessionCollector(context, options, dependencies)),
-    capture("project-harness", () => projectCollector(context, options, dependencies)),
-    capture("agent-customize", () => customizeCollector(context, options, dependencies)),
-    collectLead(context, options, leadAnalyzer),
+  const populationCollector = dependencies.collectSessionPopulation ?? collectSessionPopulation;
+  const projectPromise = capture("project-harness", () => projectCollector(context, options, dependencies));
+  const customizePromise = capture("agent-customize", () => customizeCollector(context, options, dependencies));
+  const populationLane = await capture("session-population", async () =>
+    availableLane(await populationCollector(context, options, dependencies)));
+  const sessionPopulation = laneIsAvailable(populationLane) ? populationLane.data : null;
+  const sessionPromise = sessionPopulation
+    ? capture("session-evidence", () => sessionCollector(context, options, {
+        ...dependencies,
+        sessionPopulation,
+      }))
+    : Promise.resolve(unavailableLane("session-evidence", populationLane.error));
+  const leadPromise = sessionPopulation
+    ? collectLead(context, options, leadAnalyzer, sessionPopulation)
+    : Promise.resolve(unavailableLane("lead-analyzer", populationLane.error));
+  let [sessionEvidence, projectHarness, agentCustomize, lead] = await Promise.all([
+    sessionPromise,
+    projectPromise,
+    customizePromise,
+    leadPromise,
   ]);
+  const sessionPopulationBinding = populationDiagnostics(sessionPopulation, sessionEvidence, lead);
+  if (sessionPopulationBinding.status === "conflict") {
+    lead = unavailableLane("lead-analyzer", Object.assign(
+      new Error("Session population binding mismatch"),
+      { code: "SESSION_POPULATION_BINDING_MISMATCH" },
+    ));
+  }
   const lanes = { sessionEvidence, projectHarness, agentCustomize };
   const incompleteLanes = EVIDENCE_LANE_NAMES.filter((name) => !laneIsAvailable(lanes[name]));
   const unavailableLanes = EVIDENCE_LANE_NAMES.filter((name) => lanes[name]?.status === "unavailable");
@@ -108,6 +183,7 @@ export async function collectEvidenceBundle(options = {}, dependencies = {}) {
       topologyStatus: context.topology.status,
       topologyIncomplete,
       individualCommandsRemainDiagnostic: true,
+      sessionPopulationBinding,
     },
   };
 }
diff --git a/scripts/harness-analysis/evidence-bundle/session-evidence.mjs b/scripts/harness-analysis/evidence-bundle/session-evidence.mjs
index 9047d72..71681c8 100644
--- a/scripts/harness-analysis/evidence-bundle/session-evidence.mjs
+++ b/scripts/harness-analysis/evidence-bundle/session-evidence.mjs
@@ -1,4 +1,9 @@
 import { createAnalyzer } from "../../session-analysis.mjs";
+import {
+  bindSessionSelection,
+  freezeSessionPopulation,
+  sessionAdmissionBinding,
+} from "../../session-analysis/session-population.mjs";
 import { availableLane } from "./contract.mjs";
 
 const SOURCE_COVERAGE_STATES = new Set([
@@ -9,21 +14,60 @@ const SOURCE_COVERAGE_STATES = new Set([
   "observed",
 ]);
 
+function analyzerOptions(context, options = {}) {
+  return {
+    workspace: context.workspace,
+    platform: context.provider,
+    since: context.window.since,
+    until: context.window.until,
+    ...(options[`${context.provider}-home`] ? { [`${context.provider}-home`]: options[`${context.provider}-home`] } : {}),
+  };
+}
+
+export async function collectSessionPopulation(context, options = {}, dependencies = {}) {
+  const create = dependencies.createAnalyzer ?? createAnalyzer;
+  const analyzer = await create(context.provider);
+  const runOptions = analyzerOptions(context, options);
+  const discovery = await analyzer.analyze({ ...runOptions, command: "sources" });
+  if (!Array.isArray(discovery?.sessions)) {
+    throw Object.assign(new Error("Session population discovery returned an invalid contract"), {
+      code: "INVALID_SESSION_POPULATION",
+    });
+  }
+  const scope = {
+    ...(discovery.scope ?? {}),
+    platform: context.provider,
+    workspace: context.workspace,
+    since: context.window.since,
+    until: context.window.until,
+  };
+  return freezeSessionPopulation({
+    scope,
+    sessions: discovery.sessions,
+    discovery: {
+      scope,
+      sources: discovery.sources,
+      warnings: discovery.warnings,
+    },
+    excludedSessionId: options["exclude-session-id"] ?? options.excludeSessionId,
+    providerSessionId: typeof analyzer.currentSessionId === "function" ? analyzer.currentSessionId() : null,
+    suppliedUntil: true,
+  });
+}
+
 export async function collectSessionEvidence(context, options = {}, dependencies = {}) {
   const create = dependencies.createAnalyzer ?? createAnalyzer;
   const analyzer = await create(context.provider);
+  const population = dependencies.sessionPopulation ?? null;
   const data = await analyzer.analyze({
     command: "facts",
-    workspace: context.workspace,
-    platform: context.provider,
+    ...analyzerOptions(context, options),
     selection: "all-eligible",
-    since: context.window.since,
-    until: context.window.until,
     limit: context.evidenceLimit,
     "episode-limit": context.evidenceLimit,
+    ...(population ? { sessionInventory: population.sessions } : {}),
     topology: context.topology,
     analysisScope: context.analysisScope,
-    ...(options[`${context.provider}-home`] ? { [`${context.provider}-home`]: options[`${context.provider}-home`] } : {}),
   });
   if (data?.kind !== "session-core-facts" || !Array.isArray(data.candidates)) {
     throw Object.assign(new Error("session facts returned an invalid contract"), {
@@ -39,5 +83,29 @@ export async function collectSessionEvidence(context, options = {}, dependencies
   const laneStatus = ["unobserved", "partial"].includes(sourceCoverageStatus)
     ? "partial"
     : "available";
-  return availableLane(data, laneStatus);
+  if (!population) return availableLane(data, laneStatus);
+  const eligibleCount = Number(data.scope?.eligibleSessions ?? -1);
+  const selectedCount = Number(data.scope?.selectedSessions ?? -1);
+  if (eligibleCount !== population.binding.eligible.count
+    || selectedCount !== population.binding.eligible.count) {
+    throw Object.assign(new Error("Session facts do not match the frozen all-eligible population"), {
+      code: "SESSION_POPULATION_BINDING_MISMATCH",
+    });
+  }
+  const selectionBinding = bindSessionSelection(population, population.sessions, {
+    strategy: "all-eligible",
+    projectionPolicy: "session-fact-candidates-v2",
+  });
+  const boundData = {
+    ...data,
+    omitted: {
+      ...(data.omitted ?? {}),
+      activeSessions: population.binding.omission.activeSessions,
+      homeSessionOnly: population.binding.omission.homeSessionOnly,
+    },
+    populationBinding: population.binding,
+    selectionBinding,
+  };
+  boundData.admissionBinding = sessionAdmissionBinding(boundData, selectionBinding);
+  return availableLane(boundData, laneStatus);
 }
diff --git a/scripts/harness-analysis/renderers/html.mjs b/scripts/harness-analysis/renderers/html.mjs
index a98000a..482a3ee 100644
--- a/scripts/harness-analysis/renderers/html.mjs
+++ b/scripts/harness-analysis/renderers/html.mjs
@@ -11,6 +11,24 @@ function escapeHtml(value) {
     .replace(/'/gu, "&#39;");
 }
 
+const CJK_PHRASE = /^\p{Script=Han}{2,8}$/u;
+
+function renderVisibleText(value, language) {
+  const text = String(value ?? "");
+  if (language !== "zh-CN" || typeof Intl.Segmenter !== "function") return escapeHtml(text);
+  try {
+    const segments = new Intl.Segmenter("zh-CN", { granularity: "word" }).segment(text);
+    return [...segments].map(({ segment, isWordLike }) => {
+      const escaped = escapeHtml(segment);
+      return isWordLike && CJK_PHRASE.test(segment)
+        ? `<span class="cjk-phrase">${escaped}</span>`
+        : escaped;
+    }).join("");
+  } catch {
+    return escapeHtml(text);
+  }
+}
+
 function list(value) {
   return Array.isArray(value) ? value : [];
 }
@@ -87,11 +105,11 @@ function stateTone(state) {
   return "muted";
 }
 
-function metric(label, value, note = "") {
+function metric(label, value, note = "", language = "en") {
   return `<div class="metric">
-    <span>${escapeHtml(label)}</span>
-    <strong>${escapeHtml(value)}</strong>
-    ${note ? `<small>${escapeHtml(note)}</small>` : ""}
+    <span>${renderVisibleText(label, language)}</span>
+    <strong>${renderVisibleText(value, language)}</strong>
+    ${note ? `<small>${renderVisibleText(note, language)}</small>` : ""}
   </div>`;
 }
 
@@ -120,11 +138,11 @@ function evidenceScoreLabel(summary, language) {
   return copy(language, "Loop Effectiveness", "循环有效性");
 }
 
-function section(id, eyebrow, title, body, note = "") {
+function section(id, eyebrow, title, body, note = "", language = "en") {
   return `<section class="section" data-section="${escapeHtml(id)}" aria-labelledby="section-${escapeHtml(id)}">
     <div class="section-heading">
-      <div><span class="eyebrow">${escapeHtml(eyebrow)}</span><h2 id="section-${escapeHtml(id)}">${escapeHtml(title)}</h2></div>
-      ${note ? `<p>${escapeHtml(note)}</p>` : ""}
+      <div><span class="eyebrow">${renderVisibleText(eyebrow, language)}</span><h2 id="section-${escapeHtml(id)}">${renderVisibleText(title, language)}</h2></div>
+      ${note ? `<p>${renderVisibleText(note, language)}</p>` : ""}
     </div>
     ${body}
   </section>`;
@@ -133,17 +151,19 @@ function section(id, eyebrow, title, body, note = "") {
 function renderDimensionGrid(summary, language) {
   const dimensions = list(summary?.dimensions);
   if (dimensions.length === 0) {
-    return `<p class="empty">${copy(language, "No reviewed dimension scores are available.", "暂无已复核的维度分数。")}</p>`;
+    return `<p class="empty">${renderVisibleText(copy(language, "No reviewed dimension scores are available.", "暂无已复核的维度分数。"), language)}</p>`;
   }
   return `<div class="dimension-grid">
     ${dimensions.map((row, index) => {
       const score = clamp(row.score);
+      const roundedScore = Math.round(score);
+      const label = row.label ?? row.id ?? `Dimension ${index + 1}`;
       const state = row.state ?? copy(language, "Reviewed", "已复核");
       return `<article class="dimension-card" data-dimension-id="${escapeHtml(row.id ?? index)}">
-        <div class="dimension-top"><span>${escapeHtml(row.label ?? row.id ?? `Dimension ${index + 1}`)}</span><span class="pill ${stateTone(state)}">${escapeHtml(state)}</span></div>
-        <div class="score-line"><strong>${Math.round(score)}</strong><span>/ 100</span></div>
-        <div class="track" aria-label="${escapeHtml(row.label ?? row.id)} ${Math.round(score)} of 100"><i style="width:${score}%"></i></div>
-        <p>${escapeHtml(row.summary ?? row.scoreReason ?? "")}</p>
+        <div class="dimension-top"><span>${renderVisibleText(label, language)}</span><span class="pill ${stateTone(state)}">${renderVisibleText(state, language)}</span></div>
+        <div class="score-line"><strong>${roundedScore}</strong><span>/ 100</span></div>
+        <div class="track" role="progressbar" aria-label="${escapeHtml(label)} ${roundedScore} of 100" aria-valuemin="0" aria-valuemax="100" aria-valuenow="${roundedScore}"><i style="width:${score}%"></i></div>
+        <p>${renderVisibleText(row.summary ?? row.scoreReason ?? "", language)}</p>
       </article>`;
     }).join("\n")}
   </div>`;
@@ -179,12 +199,12 @@ function renderActivity(summary, language) {
 
   return `<div class="activity-layout">
     <div class="activity-panel">
-      <div class="heatmap" role="img" aria-label="${copy(language, "Session activity heatmap", "会话活跃度热力图")}">${cells || `<span class="empty">${copy(language, "No dated activity was retained.", "没有保留按日期聚合的活动数据。")}</span>`}</div>
+      <div class="heatmap" role="img" aria-label="${copy(language, "Session activity heatmap", "会话活跃度热力图")}">${cells || `<span class="empty">${renderVisibleText(copy(language, "No dated activity was retained.", "没有保留按日期聚合的活动数据。"), language)}</span>`}</div>
       ${dates.length > 0 ? `<div class="heat-legend"><span>${escapeHtml(dates[0])}</span><span>${escapeHtml(dates.at(-1))}</span></div>` : ""}
     </div>
     <div class="stacked-metrics">
-      ${metric(copy(language, "Sessions reviewed", "已分析会话"), `${formatNumber(analyzed, language)} / ${formatNumber(eligible, language)}`, copy(language, "all-eligible usage census", "全量合格用量普查"))}
-      ${metric(copy(language, "Long-session leads", "长会话线索"), formatNumber(longCount, language), copy(language, `longest ${formatNumber(longest, language)} min`, `最长 ${formatNumber(longest, language)} 分钟`))}
+      ${metric(copy(language, "Sessions reviewed", "已分析会话"), `${formatNumber(analyzed, language)} / ${formatNumber(eligible, language)}`, copy(language, "all-eligible usage census", "全量合格用量普查"), language)}
+      ${metric(copy(language, "Long-session leads", "长会话线索"), formatNumber(longCount, language), copy(language, `longest ${formatNumber(longest, language)} min`, `最长 ${formatNumber(longest, language)} 分钟`), language)}
     </div>
   </div>`;
 }
@@ -197,22 +217,22 @@ function renderUsageLists(summary, language) {
   const renderRows = (rows, labelKey) => {
     const top = rows.slice().sort((a, b) => number(b.total ?? b.responseCount) - number(a.total ?? a.responseCount)).slice(0, 6);
     const max = Math.max(1, ...top.map((row) => number(row.total ?? row.responseCount)));
-    if (top.length === 0) return `<p class="empty">${copy(language, "No attributed rows were retained.", "没有保留已归属数据。")}</p>`;
+    if (top.length === 0) return `<p class="empty">${renderVisibleText(copy(language, "No attributed rows were retained.", "没有保留已归属数据。"), language)}</p>`;
     return `<div class="usage-list">${top.map((row) => {
       const value = number(row.total ?? row.responseCount);
-      return `<div class="usage-row"><span>${escapeHtml(row[labelKey] ?? row.name ?? "Unknown")}</span><i><b style="width:${clamp((value / max) * 100)}%"></b></i><strong>${formatNumber(value, language)}</strong></div>`;
+      return `<div class="usage-row"><span>${renderVisibleText(row[labelKey] ?? row.name ?? "Unknown", language)}</span><i><b style="width:${clamp((value / max) * 100)}%"></b></i><strong>${formatNumber(value, language)}</strong></div>`;
     }).join("")}</div>`;
   };
   return `<div class="two-column">
-    <article class="subpanel"><h3>${copy(language, "Model usage", "模型使用")}</h3>${renderRows(modelRows, "model")}</article>
-    <article class="subpanel"><h3>${copy(language, "Skill usage", "Skill 使用")}</h3>${renderRows(skillRows, "name")}</article>
+    <article class="subpanel"><h3>${renderVisibleText(copy(language, "Model usage", "模型使用"), language)}</h3>${renderRows(modelRows, "model")}</article>
+    <article class="subpanel"><h3>${renderVisibleText(copy(language, "Skill usage", "Skill 使用"), language)}</h3>${renderRows(skillRows, "name")}</article>
   </div>`;
 }
 
 function renderFindings(reportData, language) {
   const dimensions = new Map(list(reportData.summary?.dimensions).map((row) => [row.id, row.label ?? row.id]));
   const findings = list(reportData.findings);
-  if (findings.length === 0) return `<p class="empty">${copy(language, "No reviewed findings were retained.", "没有保留已复核 finding。")}</p>`;
+  if (findings.length === 0) return `<p class="empty">${renderVisibleText(copy(language, "No reviewed findings were retained.", "没有保留已复核 finding。"), language)}</p>`;
   return `<div class="finding-list">${findings.map((row, index) => {
     const prompt = findingPromptParts(row.aiFixPrompt, language);
     const expected = textLines(row.expectedOutput ?? row.expectedOutcome);
@@ -227,27 +247,27 @@ function renderFindings(reportData, language) {
     return `<article class="finding-card" data-finding-id="${escapeHtml(findingId)}">
       <div class="finding-card-main">
         <div class="finding-meta">
-          <span class="pill ${severityTone(row.severity)}">${escapeHtml(row.severity ?? "Low")}</span>
-          ${refs[0] ? `<span class="pill neutral">${escapeHtml(refs[0])}</span>` : ""}
+          <span class="pill ${severityTone(row.severity)}">${renderVisibleText(row.severity ?? "Low", language)}</span>
+          ${refs[0] ? `<span class="pill neutral">${renderVisibleText(refs[0], language)}</span>` : ""}
         </div>
-        <h3>${escapeHtml(title)}</h3>
-        <p class="finding-preview">${escapeHtml(row.reason ?? "")}</p>
+        <h3>${renderVisibleText(title, language)}</h3>
+        <p class="finding-preview">${renderVisibleText(row.reason ?? "", language)}</p>
       </div>
       <div class="finding-actions">
-        <button type="button" class="action-button secondary" data-copy-finding="${escapeHtml(findingId)}" aria-label="${escapeHtml(`${copyLabel}: ${title}`)}">${copyLabel}</button>
-        <button type="button" class="action-button ghost" data-view-finding-dialog="${dialogId}" aria-label="${escapeHtml(`${detailsLabel}: ${title}`)}">${detailsLabel}</button>
+        <button type="button" class="action-button secondary" data-copy-finding="${escapeHtml(findingId)}" aria-label="${escapeHtml(`${copyLabel}: ${title}`)}">${renderVisibleText(copyLabel, language)}</button>
+        <button type="button" class="action-button ghost" data-view-finding-dialog="${dialogId}" aria-label="${escapeHtml(`${detailsLabel}: ${title}`)}">${renderVisibleText(detailsLabel, language)}</button>
       </div>
       <dialog class="finding-dialog" id="${dialogId}" data-finding-dialog-id="${escapeHtml(findingId)}" aria-labelledby="${dialogTitleId}">
         <div class="dialog-heading">
-          <div><span class="label">${copy(language, "Finding details", "Finding 详情")}</span><h3 id="${dialogTitleId}">${escapeHtml(title)}</h3></div>
-          <span class="pill ${severityTone(row.severity)}">${escapeHtml(row.severity ?? "Low")}</span>
+          <div><span class="label">${renderVisibleText(copy(language, "Finding details", "Finding 详情"), language)}</span><h3 id="${dialogTitleId}">${renderVisibleText(title, language)}</h3></div>
+          <span class="pill ${severityTone(row.severity)}">${renderVisibleText(row.severity ?? "Low", language)}</span>
         </div>
-        <div class="dialog-section"><span class="label">${copy(language, "Cause", "原因")}</span><p>${escapeHtml(row.reason ?? "")}</p></div>
-        ${expected.length ? `<div class="dialog-section"><span class="label">${copy(language, "Expected Output", "预期结果")}</span><ol>${expected.map((item) => `<li>${escapeHtml(item)}</li>`).join("")}</ol></div>` : ""}
-        ${prompt.checks.length ? `<div class="dialog-section"><span class="label">${copy(language, "Acceptance Checks", "验收检查")}</span><ul>${prompt.checks.map((item) => `<li>${escapeHtml(item)}</li>`).join("")}</ul></div>` : ""}
+        <div class="dialog-section"><span class="label">${renderVisibleText(copy(language, "Cause", "原因"), language)}</span><p>${renderVisibleText(row.reason ?? "", language)}</p></div>
+        ${expected.length ? `<div class="dialog-section"><span class="label">${renderVisibleText(copy(language, "Expected Output", "预期结果"), language)}</span><ol>${expected.map((item) => `<li>${renderVisibleText(item, language)}</li>`).join("")}</ol></div>` : ""}
+        ${prompt.checks.length ? `<div class="dialog-section"><span class="label">${renderVisibleText(copy(language, "Acceptance Checks", "验收检查"), language)}</span><ul>${prompt.checks.map((item) => `<li>${renderVisibleText(item, language)}</li>`).join("")}</ul></div>` : ""}
         <div class="dialog-actions">
-          <button type="button" class="action-button secondary" data-copy-finding="${escapeHtml(findingId)}" aria-label="${escapeHtml(`${copyLabel}: ${title}`)}">${copyLabel}</button>
-          <button type="button" class="action-button ghost" data-close-finding-dialog="${dialogId}">${closeLabel}</button>
+          <button type="button" class="action-button secondary" data-copy-finding="${escapeHtml(findingId)}" aria-label="${escapeHtml(`${copyLabel}: ${title}`)}">${renderVisibleText(copyLabel, language)}</button>
+          <button type="button" class="action-button ghost" data-close-finding-dialog="${dialogId}">${renderVisibleText(closeLabel, language)}</button>
         </div>
       </dialog>
     </article>`;
@@ -258,20 +278,20 @@ function renderSuggestions(summary, language) {
   const suggestions = list(summary?.suggestions);
   if (suggestions.length === 0) return "";
   return `<div class="suggestion-block">
-    <div class="suggestion-heading"><div><span class="label">${copy(language, "Suggestions", "建议")}</span><h3>${copy(language, "Useful next experiments", "值得尝试的下一步")}</h3></div><p>${copy(language, "Advisory only · no AI Fix action", "仅供参考 · 不包含 AI 修复动作")}</p></div>
+    <div class="suggestion-heading"><div><span class="label">${renderVisibleText(copy(language, "Suggestions", "建议"), language)}</span><h3>${renderVisibleText(copy(language, "Useful next experiments", "值得尝试的下一步"), language)}</h3></div><p>${renderVisibleText(copy(language, "Advisory only · no AI Fix action", "仅供参考 · 不包含 AI 修复动作"), language)}</p></div>
     <div class="suggestion-list">${suggestions.map((row, index) => {
       const prerequisites = textLines(row.prerequisites);
       const blockedBy = textLines(row.blockedBy);
       return `<article class="suggestion" data-suggestion-id="${escapeHtml(row.id ?? index)}">
-        <div class="suggestion-top"><span class="pill accent">${escapeHtml(suggestionKindLabel(row.kind, language))}</span><span class="pill neutral">${escapeHtml(row.confidence)}</span></div>
-        <h3>${escapeHtml(row.title ?? row.id ?? copy(language, "Suggestion", "建议"))}</h3>
-        <p>${escapeHtml(row.reason)}</p>
+        <div class="suggestion-top"><span class="pill accent">${renderVisibleText(suggestionKindLabel(row.kind, language), language)}</span><span class="pill neutral">${renderVisibleText(row.confidence, language)}</span></div>
+        <h3>${renderVisibleText(row.title ?? row.id ?? copy(language, "Suggestion", "建议"), language)}</h3>
+        <p>${renderVisibleText(row.reason, language)}</p>
         <dl>
-          <div><dt>${copy(language, "Owner", "负责人")}</dt><dd>${escapeHtml(row.owner)}</dd></div>
-          <div><dt>${copy(language, "Next step", "下一步")}</dt><dd>${escapeHtml(row.nextStep)}</dd></div>
-          <div><dt>${copy(language, "Validation", "验证")}</dt><dd>${escapeHtml(row.validation)}</dd></div>
-          ${prerequisites.length ? `<div><dt>${copy(language, "Prerequisites", "前置条件")}</dt><dd>${prerequisites.map(escapeHtml).join(" · ")}</dd></div>` : ""}
-          ${blockedBy.length ? `<div><dt>${copy(language, "Blocked by", "阻塞项")}</dt><dd>${blockedBy.map(escapeHtml).join(" · ")}</dd></div>` : ""}
+          <div><dt>${renderVisibleText(copy(language, "Owner", "负责人"), language)}</dt><dd>${renderVisibleText(row.owner, language)}</dd></div>
+          <div><dt>${renderVisibleText(copy(language, "Next step", "下一步"), language)}</dt><dd>${renderVisibleText(row.nextStep, language)}</dd></div>
+          <div><dt>${renderVisibleText(copy(language, "Validation", "验证"), language)}</dt><dd>${renderVisibleText(row.validation, language)}</dd></div>
+          ${prerequisites.length ? `<div><dt>${renderVisibleText(copy(language, "Prerequisites", "前置条件"), language)}</dt><dd>${prerequisites.map((item) => renderVisibleText(item, language)).join(" · ")}</dd></div>` : ""}
+          ${blockedBy.length ? `<div><dt>${renderVisibleText(copy(language, "Blocked by", "阻塞项"), language)}</dt><dd>${blockedBy.map((item) => renderVisibleText(item, language)).join(" · ")}</dd></div>` : ""}
         </dl>
       </article>`;
     }).join("\n")}</div>
@@ -284,18 +304,18 @@ function renderCustomize(summary, language) {
   const rows = list(practice.coverageRows);
   const bySurface = new Map(rows.map((row) => [row.surface, row]));
   const surfaces = [...new Set([...inspected, ...rows.map((row) => row.surface).filter(Boolean)])];
-  if (surfaces.length === 0) return `<p class="empty">${copy(language, "No project customization inventory was retained.", "没有保留项目自定义能力清单。")}</p>`;
+  if (surfaces.length === 0) return `<p class="empty">${renderVisibleText(copy(language, "No project customization inventory was retained.", "没有保留项目自定义能力清单。"), language)}</p>`;
   return `<div class="custom-grid">${surfaces.map((surface) => {
     const row = bySurface.get(surface) ?? {};
     const scopes = list(row.scopes);
     const count = row.count;
-    return `<article><span class="custom-mark">${escapeHtml(String(surface).slice(0, 1).toUpperCase())}</span><div><h3>${escapeHtml(surface)}</h3><p>${scopes.length ? escapeHtml(scopes.join(" · ")) : copy(language, "Inspected", "已检查")}${count !== undefined ? ` · ${formatNumber(count, language)}` : ""}</p></div></article>`;
+    return `<article><span class="custom-mark">${renderVisibleText(String(surface).slice(0, 1).toUpperCase(), language)}</span><div><h3>${renderVisibleText(surface, language)}</h3><p>${scopes.length ? renderVisibleText(scopes.join(" · "), language) : renderVisibleText(copy(language, "Inspected", "已检查"), language)}${count !== undefined ? ` · ${formatNumber(count, language)}` : ""}</p></div></article>`;
   }).join("")}</div>`;
 }
 
 function renderEvidence(summary, language) {
   const boundary = summary?.evidenceBoundary ?? {};
-  const coverage = summary?.atAGlance?.coverage ?? {};
+  const coverage = boundary?.episodeCoverage ?? summary?.atAGlance?.coverage ?? {};
   const selection = boundary?.manifest?.selection ?? coverage?.selection ?? {};
   const learning = summary?.learningCapture ?? {};
   const facts = [
@@ -307,11 +327,11 @@ function renderEvidence(summary, language) {
     [copy(language, "Learning state", "学习状态"), learning.state ?? "—"],
   ];
   const gaps = textLines(boundary.sourceGaps);
-  return `<div class="evidence-grid">${facts.map(([label, value]) => `<div><span>${escapeHtml(label)}</span><strong>${escapeHtml(value)}</strong></div>`).join("")}</div>
-    ${gaps.length ? `<div class="evidence-note"><strong>${copy(language, "Source gaps", "来源缺口")}</strong><ul>${gaps.map((gap) => `<li>${escapeHtml(gap)}</li>`).join("")}</ul></div>` : ""}
-    <p class="method-note">${copy(language,
+  return `<div class="evidence-grid">${facts.map(([label, value]) => `<div><span>${renderVisibleText(label, language)}</span><strong>${renderVisibleText(value, language)}</strong></div>`).join("")}</div>
+    ${gaps.length ? `<div class="evidence-note"><strong>${renderVisibleText(copy(language, "Source gaps", "来源缺口"), language)}</strong><ul>${gaps.map((gap) => `<li>${renderVisibleText(gap, language)}</li>`).join("")}</ul></div>` : ""}
+    <p class="method-note">${renderVisibleText(copy(language,
       "Activity totals describe volume, not quality or savings. Fluency conclusions come from the reviewed task sample and remain bounded by the retained evidence.",
-      "活动总量只描述用量，不直接代表质量、效率或节省。流畅度结论来自已复核任务样本，并受已保留证据边界约束。")}</p>`;
+      "活动总量只描述用量，不直接代表质量、效率或节省。流畅度结论来自已复核任务样本，并受已保留证据边界约束。"), language)}</p>`;
 }
 
 function renderHtmlBody(reportData) {
@@ -335,24 +355,24 @@ function renderHtmlBody(reportData) {
 
   return `<header class="hero" data-section="overview">
     <div class="hero-copy">
-      <span class="eyebrow">${copy(language, "Harness Insights · Codex HTML", "Harness 洞察 · Codex HTML")}</span>
-      <h1>${escapeHtml(projectName)}</h1>
-      <p>${escapeHtml(overview)}</p>
-      <div class="hero-tags"><span class="pill accent">${escapeHtml(summary.modelId ?? "Harness")}</span><span class="pill muted">${copy(language, "Evidence-bound", "证据有界")}</span></div>
+      <span class="eyebrow">${renderVisibleText(copy(language, "Harness Insights · Codex HTML", "Harness 洞察 · Codex HTML"), language)}</span>
+      <h1>${renderVisibleText(projectName, language)}</h1>
+      <p>${renderVisibleText(overview, language)}</p>
+      <div class="hero-tags"><span class="pill accent">${renderVisibleText(summary.modelId ?? "Harness", language)}</span><span class="pill muted">${renderVisibleText(copy(language, "Evidence-bound", "证据有界"), language)}</span></div>
     </div>
-    <div class="score-orbit"><strong>${dimensions.length}</strong><span>${copy(language, "reviewed dimensions", "独立复核维度")}</span></div>
+    <div class="score-orbit"><strong>${dimensions.length}</strong><span>${renderVisibleText(copy(language, "reviewed dimensions", "独立复核维度"), language)}</span></div>
   </header>
   <div class="metrics">
-    ${metric(evidenceScoreLabel(summary, language), `${effectiveness} / 100`, copy(language, "Changes after later task outcomes", "等待后续任务结果后更新"))}
-    ${metric(copy(language, "Asset Health / Repair Progress", "资产健康 / 修复进度"), `${progress.score} / 100`, copy(language, `${progress.verified} verified · ${progress.partial} partial · ${progress.pending} pending`, `${progress.verified} 项已验证 · ${progress.partial} 项部分完成 · ${progress.pending} 项待处理`))}
-    ${metric(copy(language, "Sessions analyzed", "会话样本"), `${formatNumber(analyzed, language)} / ${formatNumber(eligible, language)}`, selection.confidence ?? "")}
-    ${metric(copy(language, "Findings", "待处理 Finding"), findings.length, `${high} High · ${medium} Medium`)}
+    ${metric(evidenceScoreLabel(summary, language), `${effectiveness} / 100`, copy(language, "Changes after later task outcomes", "等待后续任务结果后更新"), language)}
+    ${metric(copy(language, "Asset Health / Repair Progress", "资产健康 / 修复进度"), `${progress.score} / 100`, copy(language, `${progress.verified} verified · ${progress.partial} partial · ${progress.pending} pending`, `${progress.verified} 项已验证 · ${progress.partial} 项部分完成 · ${progress.pending} 项待处理`), language)}
+    ${metric(copy(language, "Sessions analyzed", "会话样本"), `${formatNumber(analyzed, language)} / ${formatNumber(eligible, language)}`, selection.confidence ?? "", language)}
+    ${metric(copy(language, "Findings", "待处理 Finding"), findings.length, `${high} High · ${medium} Medium`, language)}
   </div>
-  ${section("fluency", copy(language, "01 · Readiness", "01 · 就绪度"), copy(language, "Five-dimension fluency", "五维流畅度"), renderDimensionGrid(summary, language), copy(language, "Scores and states come from the reviewed source.", "分数与状态来自已复核 source。"))}
-  ${hasUsage ? section("activity", copy(language, "02 · Signals", "02 · 信号"), copy(language, "Project usage", "项目用量"), `${renderActivity(summary, language)}${renderUsageLists(summary, language)}`, copy(language, "Volume is context, not an outcome claim.", "用量是背景，不是结果结论。")) : ""}
-  ${section("findings", copy(language, "03 · Action", "03 · 行动"), copy(language, "Findings and recommendations", "Finding 与建议"), `${renderFindings(reportData, language)}${suggestions.length ? renderSuggestions(summary, language) : ""}`, taskLoopReport ? copy(language, `${findings.length} findings · ${suggestions.length} suggestions`, `${findings.length} 条 finding · ${suggestions.length} 条建议`) : copy(language, `${findings.length} reviewed items`, `${findings.length} 条已复核项`))}
-  ${section("customize", copy(language, "04 · Capability", "04 · 能力"), copy(language, "Agent Customize", "Agent 自定义"), renderCustomize(summary, language), copy(language, "Inspected project and authorized host surfaces.", "已检查的项目与授权宿主能力面。"))}
-  ${section("methodology", copy(language, "05 · Boundary", "05 · 边界"), copy(language, "Evidence and methodology", "证据与方法"), renderEvidence(summary, language), copy(language, "Reader-safe evidence only.", "仅使用读者安全证据。"))}`;
+  ${section("fluency", copy(language, "01 · Readiness", "01 · 就绪度"), copy(language, "Five-dimension fluency", "五维流畅度"), renderDimensionGrid(summary, language), copy(language, "Scores and states come from the reviewed source.", "分数与状态来自已复核 source。"), language)}
+  ${hasUsage ? section("activity", copy(language, "02 · Signals", "02 · 信号"), copy(language, "Project usage", "项目用量"), `${renderActivity(summary, language)}${renderUsageLists(summary, language)}`, copy(language, "Volume is context, not an outcome claim.", "用量是背景，不是结果结论。"), language) : ""}
+  ${section("findings", copy(language, "03 · Action", "03 · 行动"), copy(language, "Findings and recommendations", "Finding 与建议"), `${renderFindings(reportData, language)}${suggestions.length ? renderSuggestions(summary, language) : ""}`, taskLoopReport ? copy(language, `${findings.length} findings · ${suggestions.length} suggestions`, `${findings.length} 条 finding · ${suggestions.length} 条建议`) : copy(language, `${findings.length} reviewed items`, `${findings.length} 条已复核项`), language)}
+  ${section("customize", copy(language, "04 · Capability", "04 · 能力"), copy(language, "Agent Customize", "Agent 自定义"), renderCustomize(summary, language), copy(language, "Inspected project and authorized host surfaces.", "已检查的项目与授权宿主能力面。"), language)}
+  ${section("methodology", copy(language, "05 · Boundary", "05 · 边界"), copy(language, "Evidence and methodology", "证据与方法"), renderEvidence(summary, language), copy(language, "Reader-safe evidence only.", "仅使用读者安全证据。"), language)}`;
 }
 
 export function renderHtml(reportData) {
@@ -369,6 +389,7 @@ export function renderHtml(reportData) {
     :root { color-scheme: dark; --bg:#111315; --panel:#1a1d20; --panel-2:#21252a; --line:#31363d; --text:#f5f7fa; --muted:#9aa3ad; --blue:#6cb8ff; --blue-2:#1b6ca8; --orange:#ff9a5a; --green:#53d69b; --red:#ff6b6b; }
     * { box-sizing:border-box; }
     html { background:var(--bg); scroll-behavior:smooth; }
+    .cjk-phrase { white-space:nowrap; }
     body { margin:0; min-width:320px; font-family:Inter,-apple-system,BlinkMacSystemFont,"Segoe UI",sans-serif; color:var(--text); background:radial-gradient(circle at 12% -10%,#22364a 0,transparent 34rem),var(--bg); line-height:1.55; }
     .sr-only { position:absolute!important; width:1px!important; height:1px!important; padding:0!important; margin:-1px!important; overflow:hidden!important; clip:rect(0,0,0,0)!important; white-space:nowrap!important; border:0!important; }
     body::before { content:""; position:fixed; inset:0; pointer-events:none; opacity:.22; background-image:linear-gradient(rgba(255,255,255,.025) 1px,transparent 1px),linear-gradient(90deg,rgba(255,255,255,.025) 1px,transparent 1px); background-size:28px 28px; }
@@ -381,12 +402,12 @@ export function renderHtml(reportData) {
     .hero-tags { display:flex; flex-wrap:wrap; gap:8px; }
     .score-orbit { width:174px; aspect-ratio:1; display:grid; place-content:center; text-align:center; border:14px solid #343a41; border-top-color:var(--blue); border-right-color:var(--blue-2); border-radius:50%; background:#1a2026; box-shadow:inset 0 0 0 1px var(--line); }
     .score-orbit strong { font-size:44px; line-height:1; }
-    .score-orbit span { width:110px; margin-top:7px; color:var(--muted); font-size:12px; }
+    .score-orbit > span { width:110px; margin-top:7px; color:var(--muted); font-size:12px; }
     .pill { display:inline-flex; align-items:center; min-height:26px; padding:3px 10px; border-radius:999px; color:#d4dae0; background:#30353b; font-size:12px; font-weight:700; }
     .pill.accent { color:#c9e8ff; background:#164e75; }.pill.good { color:#c9f9e5; background:#165c43; }.pill.warning { color:#ffe1c8; background:#754322; }.pill.danger { color:#ffd4d4; background:#762e31; }.pill.muted,.pill.neutral { color:#c8cdd3; background:#34383d; }
     .metrics { display:grid; grid-template-columns:repeat(4,1fr); gap:14px; margin:18px 0 54px; }
     .metric { min-height:126px; padding:22px 24px; border:1px solid var(--line); border-radius:20px; background:rgba(30,34,38,.9); }
-    .metric span,.metric small { display:block; color:var(--muted); }.metric strong { display:block; margin:5px 0 2px; font-size:34px; line-height:1.1; }.metric small { font-size:12px; }
+    .metric > span,.metric > small { display:block; color:var(--muted); }.metric > strong { display:block; margin:5px 0 2px; font-size:34px; line-height:1.1; }.metric > small { font-size:12px; }
     .section { margin-top:64px; scroll-margin-top:24px; }.section-heading { display:flex; justify-content:space-between; gap:30px; align-items:end; margin-bottom:20px; }.section-heading h2 { margin:5px 0 0; font-size:30px; letter-spacing:-.025em; }.section-heading > p { max-width:420px; margin:0; color:var(--muted); text-align:right; font-size:14px; }
     .dimension-grid { display:grid; grid-template-columns:repeat(5,minmax(0,1fr)); gap:12px; }.dimension-card { min-height:230px; padding:19px; border:1px solid var(--line); border-radius:19px; background:linear-gradient(180deg,var(--panel-2),var(--panel)); }.dimension-top { display:flex; min-height:52px; justify-content:space-between; gap:8px; align-items:flex-start; font-weight:700; }.score-line { display:flex; align-items:baseline; gap:4px; margin-top:14px; }.score-line strong { font-size:34px; }.score-line span { color:var(--muted); font-size:12px; }.track { height:8px; margin:8px 0 16px; overflow:hidden; border-radius:999px; background:#30353a; }.track i { display:block; height:100%; border-radius:inherit; background:linear-gradient(90deg,var(--blue-2),var(--blue)); }.dimension-card p { margin:0; color:var(--muted); font-size:13px; }
     .activity-layout { display:grid; grid-template-columns:minmax(0,1.8fr) minmax(230px,.7fr); gap:14px; }.activity-panel,.subpanel,.finding,.custom-grid article,.evidence-grid,.evidence-note { border:1px solid var(--line); border-radius:19px; background:var(--panel); }.activity-panel { padding:22px; overflow:hidden; }.heatmap { display:grid; grid-template-rows:repeat(7,13px); grid-auto-flow:column; grid-auto-columns:13px; gap:4px; min-height:115px; overflow-x:auto; padding-bottom:8px; }.heat-cell { border-radius:3px; background:#2b3035; }.heat-cell.l1{background:#174d70}.heat-cell.l2{background:#206f9e}.heat-cell.l3{background:#319bd1}.heat-cell.l4{background:#70c9ff}.heat-legend { display:flex; justify-content:space-between; color:var(--muted); font-size:11px; }.stacked-metrics { display:grid; gap:14px; }.stacked-metrics .metric { min-height:0; }
@@ -394,7 +415,7 @@ export function renderHtml(reportData) {
     .finding-list { display:grid; grid-template-columns:repeat(3,minmax(0,1fr)); gap:12px; }.finding-card { min-width:0; min-height:220px; display:flex; flex-direction:column; padding:18px; border:1px solid var(--line); border-radius:19px; background:linear-gradient(180deg,#1c2329,var(--panel)); }.finding-card-main { display:grid; gap:10px; }.finding-meta { display:flex; flex-wrap:wrap; gap:7px; }.finding-card h3 { min-height:2.8em; margin:0; display:-webkit-box; overflow:hidden; font-size:17px; line-height:1.4; -webkit-box-orient:vertical; -webkit-line-clamp:2; }.finding-preview { margin:0; color:#c7ced6; display:-webkit-box; overflow:hidden; font-size:13px; line-height:1.55; -webkit-box-orient:vertical; -webkit-line-clamp:2; }.finding-actions,.dialog-actions { display:flex; flex-wrap:wrap; justify-content:space-between; gap:10px; margin-top:auto; padding-top:16px; }.action-button { min-height:38px; padding:8px 13px; border:1px solid transparent; border-radius:10px; color:var(--text); background:#30363d; font:inherit; font-size:13px; font-weight:700; cursor:pointer; }.action-button:hover { filter:brightness(1.12); }.action-button:focus-visible { outline:2px solid var(--blue); outline-offset:2px; }.action-button.secondary { color:#cceaff; border-color:#286d9b; background:#164e75; }.action-button.ghost { color:#d4dae0; border-color:var(--line); background:transparent; }.finding-dialog,.manual-copy-dialog { width:min(880px,calc(100vw - 32px)); max-height:calc(100vh - 32px); overflow:auto; padding:24px; border:1px solid var(--line); border-radius:20px; color:var(--text); background:var(--panel); box-shadow:0 30px 90px rgba(0,0,0,.55); }.finding-dialog::backdrop,.manual-copy-dialog::backdrop { background:rgba(4,8,12,.72); }.dialog-heading { display:flex; justify-content:space-between; gap:18px; align-items:flex-start; }.dialog-heading h3 { margin:6px 0 0; font-size:24px; }.dialog-section { margin-top:18px; }.dialog-section p { margin:7px 0 0; color:#d1d7de; }.dialog-section ol,.dialog-section ul { margin:8px 0 0; padding-left:22px; color:#d1d7de; }.manual-copy-dialog textarea { width:100%; min-height:260px; margin-top:14px; padding:14px; resize:vertical; border:1px solid var(--line); border-radius:12px; color:var(--text); background:#11161b; font:13px/1.5 ui-monospace,SFMono-Regular,Consolas,monospace; }.no-js .finding-actions,.no-js .manual-copy-dialog { display:none!important; }.no-js .finding-dialog:not([open]) { display:block; position:static; width:auto; max-height:none; margin-top:16px; box-shadow:none; }
     .suggestion-block { margin-top:28px; padding-top:24px; border-top:1px solid var(--line); }.suggestion-heading { display:flex; justify-content:space-between; gap:20px; align-items:end; margin-bottom:14px; }.suggestion-heading h3 { margin:4px 0 0; font-size:20px; }.suggestion-heading p { margin:0; color:var(--muted); font-size:12px; }.suggestion-list { display:grid; grid-template-columns:repeat(auto-fit,minmax(260px,1fr)); gap:12px; }.suggestion { padding:18px; border:1px solid var(--line); border-radius:17px; background:linear-gradient(180deg,#1b252d,var(--panel)); }.suggestion-top { display:flex; justify-content:space-between; gap:8px; }.suggestion h3 { margin:13px 0 7px; font-size:16px; }.suggestion > p { color:#c7ced6; font-size:13px; }.suggestion dl { display:grid; gap:9px; margin:14px 0 0; }.suggestion dl div { display:grid; gap:2px; }.suggestion dt { color:var(--muted); font-size:11px; font-weight:700; letter-spacing:.08em; text-transform:uppercase; }.suggestion dd { margin:0; color:#d7dde4; font-size:12px; }
     .custom-grid { display:grid; grid-template-columns:repeat(auto-fit,minmax(190px,1fr)); gap:12px; }.custom-grid article { display:flex; gap:14px; align-items:center; padding:18px; }.custom-grid h3 { margin:0; font-size:15px; }.custom-grid p { margin:3px 0 0; color:var(--muted); font-size:12px; }.custom-mark { display:grid; flex:0 0 38px; aspect-ratio:1; place-content:center; border-radius:12px; color:#d9efff; background:#164e75; font-weight:800; }
-    .evidence-grid { display:grid; grid-template-columns:repeat(3,1fr); overflow:hidden; }.evidence-grid div { padding:20px; border-right:1px solid var(--line); border-bottom:1px solid var(--line); }.evidence-grid span,.evidence-grid strong { display:block; }.evidence-grid span { color:var(--muted); font-size:12px; }.evidence-grid strong { margin-top:4px; }.evidence-note { margin-top:14px; padding:20px; }.method-note,.empty { color:var(--muted); }.method-note { max-width:800px; margin:18px 0 0; font-size:13px; }
+    .evidence-grid { display:grid; grid-template-columns:repeat(3,1fr); overflow:hidden; }.evidence-grid div { padding:20px; border-right:1px solid var(--line); border-bottom:1px solid var(--line); }.evidence-grid > div > span,.evidence-grid > div > strong { display:block; }.evidence-grid > div > span { color:var(--muted); font-size:12px; }.evidence-grid > div > strong { margin-top:4px; }.evidence-note { margin-top:14px; padding:20px; }.method-note,.empty { color:var(--muted); }.method-note { max-width:800px; margin:18px 0 0; font-size:13px; }
     footer { margin-top:64px; padding-top:20px; border-top:1px solid var(--line); color:var(--muted); font-size:12px; }
     @media (max-width:1000px) { .finding-list{grid-template-columns:repeat(2,minmax(0,1fr))} }
     @media (max-width:900px) { .metrics,.dimension-grid{grid-template-columns:repeat(2,1fr)}.activity-layout,.two-column{grid-template-columns:1fr}.evidence-grid{grid-template-columns:repeat(2,1fr)} }
@@ -407,19 +428,19 @@ export function renderHtml(reportData) {
 <div id="copy-status" class="sr-only" role="status" aria-live="polite"></div>
 <main id="harness-report" data-report-mode="codex-html">
 ${renderHtmlBody(reportData)}
-<footer>${copy(reportData.language, "Generated from one reviewed Harness source · self-contained Codex HTML", "由同一份已复核 Harness source 生成 · 自包含 Codex HTML")}</footer>
+<footer>${renderVisibleText(copy(reportData.language, "Generated from one reviewed Harness source · self-contained Codex HTML", "由同一份已复核 Harness source 生成 · 自包含 Codex HTML"), language)}</footer>
 </main>
 <dialog id="manual-copy-dialog" class="manual-copy-dialog" aria-labelledby="manual-copy-title" aria-describedby="manual-copy-description">
   <div class="dialog-heading">
     <div>
-      <span class="label">${copy(reportData.language, "Manual fallback", "手动复制")}</span>
-      <h3 id="manual-copy-title">${copy(reportData.language, "Copy the AI Fix prompt", "复制 AI 修复提示词")}</h3>
+      <span class="label">${renderVisibleText(copy(reportData.language, "Manual fallback", "手动复制"), language)}</span>
+      <h3 id="manual-copy-title">${renderVisibleText(copy(reportData.language, "Copy the AI Fix prompt", "复制 AI 修复提示词"), language)}</h3>
     </div>
   </div>
-  <p id="manual-copy-description">${copy(reportData.language, "Automatic copy was blocked. The exact prompt is selected below.", "自动复制被阻止，下面已选中完整提示词。")}</p>
+  <p id="manual-copy-description">${renderVisibleText(copy(reportData.language, "Automatic copy was blocked. The exact prompt is selected below.", "自动复制被阻止，下面已选中完整提示词。"), language)}</p>
   <textarea id="manual-copy-text" readonly aria-label="${copy(reportData.language, "AI Fix prompt", "AI 修复提示词")}"></textarea>
   <div class="dialog-actions">
-    <button type="button" class="action-button ghost" data-close-finding-dialog="manual-copy-dialog">${copy(reportData.language, "Close", "关闭")}</button>
+    <button type="button" class="action-button ghost" data-close-finding-dialog="manual-copy-dialog">${renderVisibleText(copy(reportData.language, "Close", "关闭"), language)}</button>
   </div>
 </dialog>
 <script id="harness-report-data" type="application/json">${serializeForScript(reportData)}</script>
@@ -516,8 +537,47 @@ export function evaluateHtmlReport(htmlText, reportData) {
     errors.push(`report.html suggestion row count ${suggestionRows} does not match reviewed suggestions ${list(reportData?.summary?.suggestions).length}`);
   }
   const dimensionRows = (text.match(/data-dimension-id=/gu) ?? []).length;
-  if (dimensionRows !== list(reportData?.summary?.dimensions).length) {
-    errors.push(`report.html dimension row count ${dimensionRows} does not match reviewed dimensions ${list(reportData?.summary?.dimensions).length}`);
+  const dimensions = list(reportData?.summary?.dimensions);
+  if (dimensionRows !== dimensions.length) {
+    errors.push(`report.html dimension row count ${dimensionRows} does not match reviewed dimensions ${dimensions.length}`);
+  }
+  const dimensionTracks = text.match(/<div class="track"[^>]*>/gu) ?? [];
+  if (dimensionTracks.length !== dimensions.length) {
+    errors.push(`report.html dimension progressbar count ${dimensionTracks.length} does not match reviewed dimensions ${dimensions.length}`);
+  }
+  for (const [index, dimension] of dimensions.entries()) {
+    const dimensionId = escapeHtml(dimension?.id ?? index);
+    const dimensionMarker = `data-dimension-id="${dimensionId}"`;
+    const markerCount = text.split(dimensionMarker).length - 1;
+    if (markerCount !== 1) {
+      errors.push(`report.html dimension ${dimensionId} card binding count ${markerCount} does not match expected 1`);
+      continue;
+    }
+    const cardStart = text.indexOf(dimensionMarker);
+    const cardEnd = text.indexOf("</article>", cardStart);
+    const cardText = cardEnd < 0 ? "" : text.slice(cardStart, cardEnd);
+    const cardTracks = cardText.match(/<div class="track"[^>]*>/gu) ?? [];
+    if (cardTracks.length !== 1) {
+      errors.push(`report.html dimension ${dimensionId} progressbar count ${cardTracks.length} does not match expected 1`);
+      continue;
+    }
+    const attributes = Object.fromEntries(
+      [...cardTracks[0].matchAll(/([\w-]+)="([^"]*)"/gu)].map((match) => [match[1], match[2]]),
+    );
+    const roundedScore = Math.round(clamp(dimension?.score));
+    const label = dimension?.label ?? dimension?.id ?? `Dimension ${index + 1}`;
+    const expectedAttributes = {
+      role: "progressbar",
+      "aria-label": `${escapeHtml(label)} ${roundedScore} of 100`,
+      "aria-valuemin": "0",
+      "aria-valuemax": "100",
+      "aria-valuenow": String(roundedScore),
+    };
+    for (const [attribute, expected] of Object.entries(expectedAttributes)) {
+      if (attributes[attribute] !== expected) {
+        errors.push(`report.html dimension ${dimensionId} ${attribute} ${attributes[attribute] ?? "missing"} does not match expected ${expected}`);
+      }
+    }
   }
   if (text.length < 5000) warnings.push("report.html is unexpectedly small for the self-contained visual contract");
   return {
diff --git a/scripts/harness-analysis/report-run.mjs b/scripts/harness-analysis/report-run.mjs
index 2dbcd02..607876f 100644
--- a/scripts/harness-analysis/report-run.mjs
+++ b/scripts/harness-analysis/report-run.mjs
@@ -152,9 +152,9 @@ export async function analyzeHarnessEvidence(options = {}) {
       code: "CANVAS_REPLACE_REQUIRES_OUTPUT",
     });
   }
-  const source = options.sourceInput
-    ? clone(options.sourceInput)
-    : (await createTaskLoopSourceFromSessions({
+  const sourceResult = options.sourceInput
+    ? { source: clone(options.sourceInput), sessionBinding: options.sessionBinding ?? null }
+    : await createTaskLoopSourceFromSessions({
         workspace,
         platform,
         language,
@@ -170,10 +170,12 @@ export async function analyzeHarnessEvidence(options = {}) {
         qwenHome: options["qwen-home"],
         copilotHome: options["copilot-home"],
         piHome: options["pi-home"],
+        sessionPopulation: options.sessionPopulation,
         workbuddyHome: options["workbuddy-home"],
         topology: options.topology,
         analysisScope: options.analysisScope,
-      })).source;
+      });
+  const source = sourceResult.source;
   const sourceErrors = validateHarnessReportSource(source);
   if (sourceErrors.length > 0) {
     throw Object.assign(new Error(sourceErrors.join("; ")), {
@@ -198,6 +200,7 @@ export async function analyzeHarnessEvidence(options = {}) {
     schemaVersion: HARNESS_ANALYSIS_EVIDENCE_SCHEMA_VERSION,
     evidence,
     summaryFacts,
+    ...(sourceResult.sessionBinding ? { sessionBinding: sourceResult.sessionBinding } : {}),
     ...(canvasPath ? { canvasPath, canvasReplaced } : {}),
   };
 }
diff --git a/scripts/harness-analysis/task-loop-source.mjs b/scripts/harness-analysis/task-loop-source.mjs
index 81ab835..867d084 100644
--- a/scripts/harness-analysis/task-loop-source.mjs
+++ b/scripts/harness-analysis/task-loop-source.mjs
@@ -22,6 +22,11 @@ import { buildTaskEpisodes, stableFingerprint } from "../session-analysis/episod
 import { buildObservationManifest } from "../session-analysis/observation-manifest.mjs";
 import { sanitizePrivateReviewText } from "../session-analysis/privacy-safe-text.mjs";
 import { sessionAnalysisRef } from "../session-analysis/session-ref.mjs";
+import {
+  bindSessionSelection,
+  leadAdmissionBinding,
+  sessionPopulationDiscovery,
+} from "../session-analysis/session-population.mjs";
 import { selectSessions } from "../session-analysis/selection.mjs";
 import {
   assertSessionSelectionBinding,
@@ -1092,8 +1097,12 @@ export async function createTaskLoopSourceFromSessions(options = {}) {
     topology: options.topology,
     analysisScope: options.analysisScope,
   };
-  const discovery = await analyzer.analyze({ ...analyzerOptions, command: "sources" });
-  const sessionInventory = Object.freeze(discovery.sessions.map((session) => Object.freeze(structuredClone(session))));
+  const population = options.sessionPopulation ?? null;
+  const discovery = population
+    ? sessionPopulationDiscovery(population)
+    : await analyzer.analyze({ ...analyzerOptions, command: "sources" });
+  const inventorySource = population?.sessions ?? discovery.sessions;
+  const sessionInventory = Object.freeze(inventorySource.map((session) => Object.freeze(structuredClone(session))));
   if (selectionProfile) {
     assertSessionSelectionBinding(selectionProfile, selectionPlan, { eligibleCount: sessionInventory.length });
   }
@@ -1234,7 +1243,24 @@ export async function createTaskLoopSourceFromSessions(options = {}) {
     memoryInventory: practiceInventory?.memories ?? { included: false, categories: [] },
   });
   assertStandardUsageComplete(source, selected, includeUsage);
-  return { source, selection: selected };
+  if (!population) return { source, selection: selected };
+  const selectionBinding = bindSessionSelection(population, selected.sessions, {
+    strategy: selected.strategy,
+    projectionPolicy: "lead-report-signal-v1",
+  });
+  const admittedEpisodes = Number(source.sessionEvents?.candidateEpisodeCount ?? 0);
+  const zeroSignalDiscardedEpisodes = Number(source.sessionEvents?.discardedEpisodeCount ?? 0);
+  const sessionBinding = {
+    population: population.binding,
+    selection: selectionBinding,
+    admission: leadAdmissionBinding({
+      projectedEpisodes: admittedEpisodes + zeroSignalDiscardedEpisodes,
+      admittedEpisodes,
+      zeroSignalDiscardedEpisodes,
+      retainedTaskEpisodes: source.taskEpisodes.length,
+    }, selectionBinding),
+  };
+  return { source, selection: selected, sessionBinding };
 }
 
 async function main(argv = process.argv.slice(2)) {
diff --git a/scripts/session-analysis/session-population.mjs b/scripts/session-analysis/session-population.mjs
new file mode 100644
index 0000000..3878cb4
--- /dev/null
+++ b/scripts/session-analysis/session-population.mjs
@@ -0,0 +1,259 @@
+import { stableFingerprint } from "./episode-contract.mjs";
+import {
+  createFactsRunContext,
+  prepareFactsSessionInventory,
+} from "./session-core-facts.mjs";
+
+export const SESSION_POPULATION_BINDING_SCHEMA_VERSION = 1;
+export const SESSION_SELECTION_BINDING_SCHEMA_VERSION = 1;
+export const SESSION_ADMISSION_BINDING_SCHEMA_VERSION = 1;
+
+const PRIVATE_POPULATION_IDS = new WeakMap();
+const PRIVATE_POPULATION_DISCOVERY = new WeakMap();
+
+function rows(value) {
+  return Array.isArray(value) ? value : [];
+}
+
+function count(value) {
+  const numeric = Number(value ?? 0);
+  return Number.isInteger(numeric) && numeric >= 0 ? numeric : 0;
+}
+
+function sessionIds(sessions) {
+  return [...new Set(rows(sessions).map((session) => String(session?.sessionId ?? "").trim()).filter(Boolean))].sort();
+}
+
+function samePopulation(left, right) {
+  return left?.schemaVersion === right?.schemaVersion
+    && left?.kind === right?.kind
+    && left?.scopeFingerprint === right?.scopeFingerprint
+    && left?.policyFingerprint === right?.policyFingerprint
+    && left?.eligible?.count === right?.eligible?.count
+    && left?.eligible?.fingerprint === right?.eligible?.fingerprint
+    && left?.omission?.exactIdentityAvailable === right?.omission?.exactIdentityAvailable
+    && left?.omission?.activeSessions === right?.omission?.activeSessions
+    && left?.omission?.homeSessionOnly === right?.omission?.homeSessionOnly
+    && left?.omission?.recencyInference === right?.omission?.recencyInference;
+}
+
+function bindingError(code, message) {
+  return Object.assign(new Error(message), { code });
+}
+
+export function freezeSessionPopulation({
+  scope = {},
+  sessions = [],
+  discovery = null,
+  excludedSessionId = null,
+  providerSessionId = null,
+  suppliedUntil = scope.until !== undefined && scope.until !== null && scope.until !== "",
+  startedAt = Date.now(),
+} = {}) {
+  const platform = String(scope.platform ?? "unknown");
+  const factsContext = createFactsRunContext({
+    workspace: scope.workspace,
+    since: scope.since,
+    ...(suppliedUntil ? { until: scope.until } : {}),
+    ...(excludedSessionId ? { "exclude-session-id": excludedSessionId } : {}),
+    _factsStartedAt: startedAt,
+  }, platform, providerSessionId);
+  const prepared = prepareFactsSessionInventory(rows(sessions), factsContext);
+  const frozenSessions = Object.freeze(prepared.sessions.map((session) => Object.freeze(structuredClone(session))));
+  const ids = sessionIds(frozenSessions);
+  const population = {
+    sessions: frozenSessions,
+    binding: Object.freeze({
+      schemaVersion: SESSION_POPULATION_BINDING_SCHEMA_VERSION,
+      kind: "session-population-binding",
+      scopeFingerprint: stableFingerprint({
+        platform,
+        workspace: scope.workspace ?? null,
+        since: scope.since ?? null,
+        until: scope.until ?? null,
+      }),
+      policyFingerprint: stableFingerprint({
+        platform,
+        qoderHomeOnlyExcluded: platform === "qoder",
+        exactIdentityAvailable: Boolean(factsContext.excludedSessionId),
+        recencyInference: suppliedUntil ? "disabled-frozen-until" : "enabled-unfrozen-until",
+      }),
+      omission: Object.freeze({
+        exactIdentityAvailable: Boolean(factsContext.excludedSessionId),
+        activeSessions: count(prepared.omitted.activeSessions),
+        homeSessionOnly: count(prepared.omitted.homeSessionOnly),
+        recencyInference: suppliedUntil ? "disabled-frozen-until" : "enabled-unfrozen-until",
+      }),
+      eligible: Object.freeze({
+        count: ids.length,
+        fingerprint: stableFingerprint(ids),
+      }),
+    }),
+  };
+  PRIVATE_POPULATION_IDS.set(population, new Set(ids));
+  PRIVATE_POPULATION_DISCOVERY.set(population, Object.freeze({
+    scope: Object.freeze(structuredClone(discovery?.scope ?? scope)),
+    sources: Object.freeze(rows(discovery?.sources).map((source) => Object.freeze(structuredClone(source)))),
+    warnings: Object.freeze(rows(discovery?.warnings).map((warning) => Object.freeze(structuredClone(warning)))),
+  }));
+  return Object.freeze(population);
+}
+
+export function sessionPopulationDiscovery(population) {
+  const discovery = PRIVATE_POPULATION_DISCOVERY.get(population);
+  if (!discovery) {
+    throw bindingError("INVALID_SESSION_POPULATION", "discovery requires a private frozen Session population");
+  }
+  return discovery;
+}
+
+export function bindSessionSelection(population, selectedSessions = [], {
+  strategy = "stratified",
+  projectionPolicy = "session-projection-v1",
+} = {}) {
+  const eligibleIds = PRIVATE_POPULATION_IDS.get(population);
+  if (!eligibleIds) {
+    throw bindingError("INVALID_SESSION_POPULATION", "selection requires a private frozen Session population");
+  }
+  const selectedIds = sessionIds(selectedSessions);
+  if (selectedIds.some((id) => !eligibleIds.has(id))) {
+    throw bindingError(
+      "SESSION_SELECTION_OUTSIDE_POPULATION",
+      "selected Session population is not a subset of the frozen eligible population",
+    );
+  }
+  return Object.freeze({
+    schemaVersion: SESSION_SELECTION_BINDING_SCHEMA_VERSION,
+    kind: "session-selection-binding",
+    parentPopulationFingerprint: population.binding.eligible.fingerprint,
+    strategy: String(strategy),
+    selected: Object.freeze({
+      count: selectedIds.length,
+      fingerprint: stableFingerprint(selectedIds),
+    }),
+    projectionPolicyFingerprint: stableFingerprint(String(projectionPolicy)),
+  });
+}
+
+export function sessionAdmissionBinding(data = {}, selection = {}) {
+  const admission = data.admission ?? {};
+  const omitted = data.omitted ?? {};
+  return Object.freeze({
+    schemaVersion: SESSION_ADMISSION_BINDING_SCHEMA_VERSION,
+    kind: "session-admission-binding",
+    projectionPolicyFingerprint: selection.projectionPolicyFingerprint,
+    taskEpisodes: count(admission.taskEpisodes),
+    candidateEpisodes: count(admission.candidateEpisodes),
+    distinctRequests: count(admission.distinctRequests),
+    emittedCandidates: count(admission.emittedCandidates),
+    noRequest: count(omitted.noRequest),
+    selfAnalysis: count(omitted.selfAnalysis),
+    lowSignal: count(omitted.lowSignal),
+    duplicateRequests: count(omitted.duplicateRequests),
+    candidateBudget: count(omitted.candidateBudget),
+  });
+}
+
+export function leadAdmissionBinding(data = {}, selection = {}) {
+  return Object.freeze({
+    schemaVersion: SESSION_ADMISSION_BINDING_SCHEMA_VERSION,
+    kind: "lead-admission-binding",
+    projectionPolicyFingerprint: selection.projectionPolicyFingerprint,
+    projectedEpisodes: count(data.projectedEpisodes),
+    admittedEpisodes: count(data.admittedEpisodes),
+    zeroSignalDiscardedEpisodes: count(data.zeroSignalDiscardedEpisodes),
+    retainedTaskEpisodes: count(data.retainedTaskEpisodes),
+  });
+}
+
+function populationErrors(expected, owner, binding) {
+  if (!binding || binding.kind !== "session-population-binding") {
+    return [`${owner} population binding is missing`];
+  }
+  return samePopulation(expected, binding) ? [] : [`${owner} population binding does not match the frozen population`];
+}
+
+function selectionErrors(population, owner, selection) {
+  const errors = [];
+  if (!selection || selection.kind !== "session-selection-binding") {
+    return [`${owner} selection binding is missing`];
+  }
+  if (selection.parentPopulationFingerprint !== population?.eligible?.fingerprint) {
+    errors.push(`${owner} selection is not bound to the frozen eligible population`);
+  }
+  if (count(selection.selected?.count) > count(population?.eligible?.count)) {
+    errors.push(`${owner} selected count exceeds the eligible population`);
+  }
+  if (selection.strategy === "all-eligible"
+    && (selection.selected?.count !== population?.eligible?.count
+      || selection.selected?.fingerprint !== population?.eligible?.fingerprint)) {
+    errors.push(`${owner} all-eligible selection does not equal the eligible population`);
+  }
+  return errors;
+}
+
+function sessionAdmissionErrors(admission) {
+  if (!admission || admission.kind !== "session-admission-binding") {
+    return ["Session admission binding is missing"];
+  }
+  const errors = [];
+  if (admission.taskEpisodes !== admission.candidateEpisodes
+    + admission.noRequest + admission.selfAnalysis + admission.lowSignal) {
+    errors.push("Session task Episode admission does not reconcile");
+  }
+  if (admission.candidateEpisodes !== admission.distinctRequests + admission.duplicateRequests) {
+    errors.push("Session candidate Episode admission does not reconcile");
+  }
+  if (admission.distinctRequests !== admission.emittedCandidates + admission.candidateBudget) {
+    errors.push("Session distinct-request admission does not reconcile");
+  }
+  return errors;
+}
+
+function leadAdmissionErrors(admission) {
+  if (!admission || admission.kind !== "lead-admission-binding") {
+    return ["lead admission binding is missing"];
+  }
+  const errors = [];
+  if (admission.projectedEpisodes !== admission.admittedEpisodes + admission.zeroSignalDiscardedEpisodes) {
+    errors.push("lead projected Episode admission does not reconcile");
+  }
+  if (admission.admittedEpisodes !== admission.retainedTaskEpisodes) {
+    errors.push("lead retained Episode admission does not reconcile");
+  }
+  return errors;
+}
+
+export function validateSessionPopulationBundle({ population, session, lead } = {}) {
+  const errors = [
+    ...populationErrors(population, "Session", session?.population),
+    ...populationErrors(population, "lead", lead?.population),
+    ...selectionErrors(population, "Session", session?.selection),
+    ...selectionErrors(population, "lead", lead?.selection),
+    ...sessionAdmissionErrors(session?.admission),
+    ...leadAdmissionErrors(lead?.admission),
+  ];
+  if (session?.admission?.projectionPolicyFingerprint !== session?.selection?.projectionPolicyFingerprint) {
+    errors.push("Session admission policy does not match its selection binding");
+  }
+  if (lead?.admission?.projectionPolicyFingerprint !== lead?.selection?.projectionPolicyFingerprint) {
+    errors.push("lead admission policy does not match its selection binding");
+  }
+  if (session?.selection?.selected?.fingerprint === lead?.selection?.selected?.fingerprint
+    && session?.selection?.projectionPolicyFingerprint === lead?.selection?.projectionPolicyFingerprint
+    && session?.admission?.taskEpisodes !== lead?.admission?.projectedEpisodes) {
+    errors.push("comparable Session and lead Episode totals do not match");
+  }
+  return errors;
+}
+
+export function assertSessionPopulationBundle(bindings) {
+  const errors = validateSessionPopulationBundle(bindings);
+  if (errors.length > 0) {
+    throw Object.assign(new Error(errors.join("; ")), {
+      code: "SESSION_POPULATION_BINDING_MISMATCH",
+      errors,
+    });
+  }
+  return bindings;
+}
diff --git a/test/agent-lint.test.mjs b/test/agent-lint.test.mjs
index 72343ff..05953ac 100644
--- a/test/agent-lint.test.mjs
+++ b/test/agent-lint.test.mjs
@@ -334,8 +334,11 @@ test("Claude and Codex instruction ledgers keep their active filenames separate"
       claude.graph.entrypoints.map((entrypoint) => entrypoint.relativePath),
       ["CLAUDE.md", ".claude/CLAUDE.md", "CLAUDE.local.md", ".claude/rules/security.md", "pkg/CLAUDE.md"],
     );
+    const claudeSources = claude.hostInstructionReview.sources;
     assert.deepEqual(
-      claude.hostInstructionReview.sources.map((source) => source.relativePath),
+      claudeSources
+        .filter((source) => source.scope !== "ancestor")
+        .map((source) => source.relativePath),
       [
         "user:CLAUDE.md",
         ".claude/CLAUDE.md",
@@ -345,7 +348,14 @@ test("Claude and Codex instruction ledgers keep their active filenames separate"
         "CLAUDE.local.md",
       ],
     );
-    assert.equal(claude.hostInstructionReview.sources.some((source) => source.relativePath.includes("AGENTS")), false);
+    assert.equal(
+      claudeSources
+        .filter((source) => source.scope === "ancestor")
+        .every((source) => source.sourceKind === "claude-ancestor"
+          && source.relativePath === "ancestor:CLAUDE.md"),
+      true,
+    );
+    assert.equal(claudeSources.some((source) => source.relativePath.includes("AGENTS")), false);
     assert.deepEqual(
       codex.graph.entrypoints.map((entrypoint) => entrypoint.relativePath),
       ["AGENTS.md", "pkg/AGENTS.md"],
diff --git a/test/better-harness-evidence-bundle.test.mjs b/test/better-harness-evidence-bundle.test.mjs
index d159c37..0b16130 100644
--- a/test/better-harness-evidence-bundle.test.mjs
+++ b/test/better-harness-evidence-bundle.test.mjs
@@ -14,6 +14,101 @@ import { EVIDENCE_BUNDLE_HELP } from "../scripts/harness-analysis/evidence-bundl
 
 const NOW = new Date("2026-07-24T08:00:00.000Z");
 
+const POPULATION_BINDING = Object.freeze({
+  schemaVersion: 1,
+  kind: "session-population-binding",
+  scopeFingerprint: "1111111111111111",
+  policyFingerprint: "2222222222222222",
+  omission: {
+    exactIdentityAvailable: true,
+    activeSessions: 1,
+    homeSessionOnly: 0,
+    recencyInference: "disabled-frozen-until",
+  },
+  eligible: { count: 1, fingerprint: "3333333333333333" },
+});
+
+const SESSION_SELECTION_BINDING = Object.freeze({
+  schemaVersion: 1,
+  kind: "session-selection-binding",
+  parentPopulationFingerprint: POPULATION_BINDING.eligible.fingerprint,
+  strategy: "all-eligible",
+  selected: { count: 1, fingerprint: "3333333333333333" },
+  projectionPolicyFingerprint: "4444444444444444",
+});
+
+const LEAD_SELECTION_BINDING = Object.freeze({
+  ...SESSION_SELECTION_BINDING,
+  strategy: "stratified",
+  projectionPolicyFingerprint: "5555555555555555",
+});
+
+function sessionFacts(overrides = {}) {
+  return {
+    kind: "session-core-facts",
+    candidates: [],
+    scope: { eligibleSessions: 1, selectedSessions: 1 },
+    admission: {
+      taskEpisodes: 1,
+      candidateEpisodes: 1,
+      distinctRequests: 1,
+      emittedCandidates: 1,
+    },
+    omitted: {
+      noRequest: 0,
+      selfAnalysis: 0,
+      lowSignal: 0,
+      duplicateRequests: 0,
+      candidateBudget: 0,
+      activeSessions: 1,
+      homeSessionOnly: 0,
+    },
+    populationBinding: POPULATION_BINDING,
+    selectionBinding: SESSION_SELECTION_BINDING,
+    admissionBinding: {
+      schemaVersion: 1,
+      kind: "session-admission-binding",
+      projectionPolicyFingerprint: SESSION_SELECTION_BINDING.projectionPolicyFingerprint,
+      taskEpisodes: 1,
+      candidateEpisodes: 1,
+      distinctRequests: 1,
+      emittedCandidates: 1,
+      noRequest: 0,
+      selfAnalysis: 0,
+      lowSignal: 0,
+      duplicateRequests: 0,
+      candidateBudget: 0,
+    },
+    ...overrides,
+  };
+}
+
+function leadEvidence(overrides = {}) {
+  return {
+    evidence: "bounded",
+    summaryFacts: {
+      evidenceBoundary: {
+        manifest: { selection: { eligibleCount: 1, analyzedCount: 1 } },
+        episodeCoverage: { episodeCount: 0 },
+      },
+    },
+    sessionBinding: {
+      population: POPULATION_BINDING,
+      selection: LEAD_SELECTION_BINDING,
+      admission: {
+        schemaVersion: 1,
+        kind: "lead-admission-binding",
+        projectedEpisodes: 1,
+        admittedEpisodes: 0,
+        zeroSignalDiscardedEpisodes: 1,
+        retainedTaskEpisodes: 0,
+        projectionPolicyFingerprint: LEAD_SELECTION_BINDING.projectionPolicyFingerprint,
+      },
+    },
+    ...overrides,
+  };
+}
+
 test("evidence-bundle help advertises WorkBuddy and its isolated home override", () => {
   assert.match(EVIDENCE_BUNDLE_HELP, /pi, or workbuddy/u);
   assert.match(EVIDENCE_BUNDLE_HELP, /--workbuddy-home <dir>/u);
@@ -54,13 +149,21 @@ function topologyResolution(workspace = ".", status = "complete") {
 }
 
 function dependencies(overrides = {}) {
+  const population = Object.freeze({
+    sessions: Object.freeze([{ sessionId: "eligible-session" }]),
+    binding: POPULATION_BINDING,
+  });
   return {
     now: () => NOW,
     resolveWorkspaceTopology: async ({ workspace }) => topologyResolution(workspace),
-    collectSessionEvidence: async () => availableLane({ kind: "session-core-facts", candidates: [] }),
+    collectSessionPopulation: async () => population,
+    collectSessionEvidence: async (_context, _options, received) => {
+      assert.equal(received.sessionPopulation, population);
+      return availableLane(sessionFacts());
+    },
     collectProjectHarness: async () => availableLane({ kind: "core-change-watch-evidence-pack" }),
     collectAgentCustomize: async () => availableLane({ kind: "agent-asset-baseline", status: "complete" }),
-    analyzeHarnessEvidence: async () => ({ evidence: "bounded", summaryFacts: { dimensions: [] } }),
+    analyzeHarnessEvidence: async () => leadEvidence(),
     ...overrides,
   };
 }
@@ -105,7 +208,7 @@ test("evidence bundle resolves topology once and shares the frozen binding with
     },
     collectSessionEvidence: async (context) => {
       received.push(context.topology);
-      return availableLane({ kind: "session-core-facts", candidates: [] });
+      return availableLane(sessionFacts());
     },
     collectProjectHarness: async (context) => {
       received.push(context.topology);
@@ -117,7 +220,7 @@ test("evidence bundle resolves topology once and shares the frozen binding with
     },
     analyzeHarnessEvidence: async (options) => {
       received.push(options.topology);
-      return { evidence: "bounded", summaryFacts: { dimensions: [] } };
+      return leadEvidence();
     },
   }));
 
@@ -354,6 +457,107 @@ test("Pi agentCustomize lane routes the provider and isolated config paths", asy
   assert.equal(received["include-user-home"], true);
 });
 
+test("shared Session population excludes the active session before both lanes hydrate", async () => {
+  const population = Object.freeze({
+    sessions: Object.freeze([{ sessionId: "eligible-session" }]),
+    binding: POPULATION_BINDING,
+  });
+  let lanePopulation;
+  let leadPopulation;
+  const result = await collectEvidenceBundle({
+    workspace: ".",
+    platform: "codex",
+    depth: "normal",
+  }, dependencies({
+    collectSessionPopulation: async () => population,
+    collectSessionEvidence: async (_context, _options, received) => {
+      lanePopulation = received.sessionPopulation;
+      return availableLane(sessionFacts());
+    },
+    analyzeHarnessEvidence: async (options) => {
+      leadPopulation = options.sessionPopulation;
+      return leadEvidence();
+    },
+  }));
+
+  assert.equal(lanePopulation, population);
+  assert.equal(leadPopulation, population);
+  assert.equal(result.status, "complete");
+  assert.equal(result.diagnostics.sessionPopulationBinding.status, "bound");
+  assert.equal(result.diagnostics.sessionPopulationBinding.population.eligible.count, 1);
+  assert.doesNotMatch(JSON.stringify(result), /eligible-session/u);
+});
+
+test("Session population conflict fails closed with a redacted stable code", async () => {
+  const result = await collectEvidenceBundle({ workspace: ".", platform: "codex" }, dependencies({
+    analyzeHarnessEvidence: async () => leadEvidence({
+      sessionBinding: {
+        ...leadEvidence().sessionBinding,
+        population: {
+          ...POPULATION_BINDING,
+          eligible: { count: 2, fingerprint: "6666666666666666" },
+        },
+      },
+    }),
+  }));
+
+  assert.equal(result.status, "failed");
+  assert.equal(result.lead.status, "unavailable");
+  assert.equal(result.lead.error.code, "SESSION_POPULATION_BINDING_MISMATCH");
+  assert.equal(result.diagnostics.sessionPopulationBinding.status, "conflict");
+  const serialized = JSON.stringify(result);
+  assert.doesNotMatch(serialized, /eligible-session/u);
+  assert.doesNotMatch(serialized, /"sessionId"/u);
+});
+
+test("Session population conflict rejects lead counts that contradict its binding", async () => {
+  const result = await collectEvidenceBundle({ workspace: ".", platform: "codex" }, dependencies({
+    analyzeHarnessEvidence: async () => leadEvidence({
+      summaryFacts: {
+        evidenceBoundary: {
+          manifest: { selection: { eligibleCount: 2, analyzedCount: 1 } },
+          episodeCoverage: { episodeCount: 0 },
+        },
+      },
+    }),
+  }));
+
+  assert.equal(result.status, "failed");
+  assert.equal(result.lead.error.code, "SESSION_POPULATION_BINDING_MISMATCH");
+  assert.equal(result.diagnostics.sessionPopulationBinding.status, "conflict");
+});
+
+test("Session facts reject counts that contradict the shared all-eligible population", async () => {
+  await assert.rejects(
+    collectSessionEvidence(freezeEvidenceBundleContext({ workspace: ".", platform: "codex" }, NOW), {}, {
+      sessionPopulation: {
+        sessions: [{ sessionId: "eligible-session" }],
+        binding: POPULATION_BINDING,
+      },
+      createAnalyzer: async () => ({
+        analyze: async () => sessionFacts({
+          scope: { eligibleSessions: 1, selectedSessions: 0 },
+        }),
+      }),
+    }),
+    (error) => error?.code === "SESSION_POPULATION_BINDING_MISMATCH",
+  );
+});
+
+test("zero-signal Episode admission remains valid inside one bound population", async () => {
+  const result = await collectEvidenceBundle({ workspace: ".", platform: "codex" }, dependencies());
+
+  assert.equal(result.status, "complete");
+  assert.equal(result.lead.status, "available");
+  assert.deepEqual(result.diagnostics.sessionPopulationBinding.episodes, {
+    comparison: "not-comparable-selection-or-policy",
+    sessionTaskEpisodes: 1,
+    leadProjectedEpisodes: 1,
+    leadRetainedEpisodes: 0,
+    leadZeroSignalDiscardedEpisodes: 1,
+  });
+});
+
 test("WorkBuddy agentCustomize lane routes the provider and isolated config paths", async () => {
   const context = freezeEvidenceBundleContext({
     workspace: ".",
diff --git a/test/core-change-watch.test.mjs b/test/core-change-watch.test.mjs
index 6cba292..0f4e8fd 100644
--- a/test/core-change-watch.test.mjs
+++ b/test/core-change-watch.test.mjs
@@ -89,6 +89,99 @@ test("project profile identifies language, manifest, source root, and entry sign
   }
 });
 
+test("framework-specific evidence distinguishes FastAPI from generic Rails and TypeScript paths", async () => {
+  const repo = await makeRepo({
+    "pyproject.toml": "[project]\nname = \"fastapi-service\"\ndependencies = [\"fastapi>=0.116\"]\n",
+    "app/api/routes.py": "def route(): return True\n",
+    "app/services/search.py": "def search(): return []\n",
+  });
+
+  try {
+    const profile = await analyzeProjectProfile({ cwd: repo });
+    const core = await analyzeCoreCandidates({ cwd: repo, profile, maxCommits: 20 });
+
+    assert.ok(profile.frameworks.some((item) => item.name === "fastapi" && item.confidence === "high"));
+    assert.equal(profile.frameworks.some((item) => item.name === "rails"), false);
+    assert.equal(
+      core.candidates
+        .filter((item) => item.path === "app/api" || item.path.startsWith("app/api/"))
+        .some((item) => item.reasons.some((reason) => /typescript/iu.test(reason))),
+      false,
+    );
+  } finally {
+    await rm(repo, { recursive: true, force: true });
+  }
+});
+
+test("Just recipes are projected statically as unverified argv entry candidates", async () => {
+  const repo = await makeRepo({
+    "Justfile": [
+      "default:",
+      "    @just --list",
+      "sync:",
+      "    uv sync",
+      "api-dev *args=\"\":",
+      "    uv run uvicorn app.main:app {{args}}",
+      "health:",
+      "    curl http://localhost:7102/health",
+      "[private]",
+      "secret-maintenance:",
+      "    echo hidden",
+      "check: sync",
+      "    just test",
+      "",
+    ].join("\r\n"),
+    "app/main.py": "def main(): return True\n",
+  });
+
+  try {
+    const profile = await analyzeProjectProfile({ cwd: repo });
+    const entries = profile.projectInfo.entryCandidates.filter((item) => item.kind === "just-recipe");
+
+    for (const recipe of ["default", "sync", "api-dev", "health", "check"]) {
+      const entry = entries.find((item) => item.command?.join(" ") === `just ${recipe}`);
+      assert.ok(entry, recipe);
+      assert.equal(entry.sourcePath, "Justfile");
+      assert.equal(entry.executionStatus, "unverified");
+    }
+    assert.equal(entries.some((item) => item.command?.includes("secret-maintenance")), false);
+    assert.deepEqual(
+      profile.manifests.find((item) => item.path === "Justfile")?.scripts,
+      ["api-dev", "check", "default", "health", "sync"],
+    );
+  } finally {
+    await rm(repo, { recursive: true, force: true });
+  }
+});
+
+test("non-current historical paths stay in history but leave current reads and actions", async () => {
+  const repo = await makeRepo({
+    "src/core/current.ts": "export const current = true;\n",
+    "src/core/legacy.ts": "export const legacy = 0;\n",
+  });
+
+  try {
+    for (let index = 1; index <= 4; index += 1) {
+      await writeFixtureFile(repo, "src/core/legacy.ts", `export const legacy = ${index};\n`);
+      git(repo, ["add", "."]);
+      git(repo, ["commit", "-q", "-m", `touch legacy ${index}`]);
+    }
+    git(repo, ["rm", "-q", "src/core/legacy.ts"]);
+    git(repo, ["commit", "-q", "-m", "remove legacy source"]);
+    await writeFixtureFile(repo, "src/core/new-service.ts", "export const added = true;\n");
+
+    const pack = await buildEvidencePack({ cwd: repo, baseRef: "HEAD", maxCommits: 30 });
+    const actionPaths = pack.followUpActions.flatMap((action) => action.files ?? []).map((item) => item.path);
+
+    assert.ok(pack.historyProfile.hotFiles.some((item) => item.path === "src/core/legacy.ts"));
+    assert.equal(pack.recommendedReads.some((item) => item.path === "src/core/legacy.ts"), false);
+    assert.equal(actionPaths.includes("src/core/legacy.ts"), false);
+    assert.ok(pack.recommendedReads.some((item) => item.path === "src/core/new-service.ts"));
+  } finally {
+    await rm(repo, { recursive: true, force: true });
+  }
+});
+
 test("project profile summarizes project identity, scale, and AGENTS instruction density", async () => {
   const repo = await makeRepo({
     "package.json": JSON.stringify({
diff --git a/test/harness-report-render-cli.test.mjs b/test/harness-report-render-cli.test.mjs
index e108036..4c76159 100644
--- a/test/harness-report-render-cli.test.mjs
+++ b/test/harness-report-render-cli.test.mjs
@@ -830,6 +830,183 @@ test("HTML mode mirrors the reviewed Agent Work Loop reader sections without Can
   });
 });
 
+test("HTML dimension progressbar semantics stay complete and score-bound", () => {
+  // Given: a canonical reviewed report with five fluency dimensions.
+  const reportData = {
+    ...sampleFindings(),
+    language: "en",
+    target: { name: "render-fixture", path: "/tmp/render-fixture" },
+  };
+
+  // When: the report is rendered and its dimension progressbars are inspected.
+  const html = renderHtml(reportData);
+  const progressbars = html.match(/<div class="track" role="progressbar"[^>]*>/gu) ?? [];
+
+  // Then: every displayed rounded score has one complete semantic contract.
+  assert.equal(progressbars.length, reportData.summary.dimensions.length);
+  for (const dimension of reportData.summary.dimensions) {
+    const score = Math.round(dimension.score);
+    assert.ok(progressbars.includes(
+      `<div class="track" role="progressbar" aria-label="${dimension.label} ${score} of 100" aria-valuemin="0" aria-valuemax="100" aria-valuenow="${score}">`,
+    ));
+  }
+  assert.equal(evaluateHtmlReport(html, reportData).status, "pass");
+
+  const firstProgressbar = progressbars[0] ?? "";
+  const firstProgressbarMarkup = html.match(/<div class="track" role="progressbar"[^>]*><i[^>]*><\/i><\/div>/u)?.[0] ?? "";
+  assert.notEqual(firstProgressbarMarkup, "");
+  const mutateFirstProgressbar = (pattern, replacement) => html.replace(
+    firstProgressbar,
+    firstProgressbar.replace(pattern, replacement),
+  );
+  const mutations = [
+    ["missing progressbar", html.replace(firstProgressbarMarkup, "")],
+    ["extra progressbar", html.replace(firstProgressbarMarkup, `${firstProgressbarMarkup}${firstProgressbarMarkup}`)],
+    ["missing role", mutateFirstProgressbar(' role="progressbar"', "")],
+    ["invalid role", mutateFirstProgressbar('role="progressbar"', 'role="meter"')],
+    ["missing label", mutateFirstProgressbar(/ aria-label="[^"]+"/u, "")],
+    ["invalid label", mutateFirstProgressbar(/aria-label="[^"]+"/u, 'aria-label=""')],
+    ["invalid minimum", mutateFirstProgressbar('aria-valuemin="0"', 'aria-valuemin="1"')],
+    ["invalid maximum", mutateFirstProgressbar('aria-valuemax="100"', 'aria-valuemax="99"')],
+    ["invalid current value", mutateFirstProgressbar(/aria-valuenow="[^"]+"/u, 'aria-valuenow="invalid"')],
+    ["score mismatch", mutateFirstProgressbar('aria-valuenow="62"', 'aria-valuenow="61"')],
+  ];
+
+  for (const [label, mutatedHtml] of mutations) {
+    assert.equal(
+      evaluateHtmlReport(mutatedHtml, reportData).status,
+      "fail",
+      `${label} mutation must fail validation`,
+    );
+  }
+});
+
+test("HTML evidence episode coverage preserves canonical summary facts and legacy fallback", () => {
+  // Given: machine-owned summary facts and a conflicting legacy projection.
+  const fixture = sampleFindings();
+  const reportData = {
+    ...fixture,
+    language: "en",
+    target: { name: "render-fixture", path: "/tmp/render-fixture" },
+    summary: {
+      ...fixture.summary,
+      evidenceBoundary: {
+        episodeCoverage: {
+          episodeCount: 14,
+          editedEpisodeCount: 12,
+        },
+      },
+      atAGlance: {
+        coverage: {
+          episodeCount: 1,
+          editedEpisodeCount: 1,
+        },
+      },
+    },
+  };
+
+  // When: canonical and legacy reports are rendered.
+  const canonicalHtml = renderHtml(reportData);
+  const legacyHtml = renderHtml({
+    ...reportData,
+    summary: {
+      ...reportData.summary,
+      evidenceBoundary: undefined,
+      atAGlance: {
+        coverage: {
+          episodeCount: 7,
+          editedEpisodeCount: 5,
+        },
+      },
+    },
+  });
+
+  // Then: machine facts win, while legacy-only input remains readable.
+  assert.match(canonicalHtml, /<span>Task episodes<\/span><strong>14<\/strong>/u);
+  assert.match(canonicalHtml, /<span>Edited episodes<\/span><strong>12<\/strong>/u);
+  assert.match(legacyHtml, /<span>Task episodes<\/span><strong>7<\/strong>/u);
+  assert.match(legacyHtml, /<span>Edited episodes<\/span><strong>5<\/strong>/u);
+});
+
+test("HTML CJK phrase breaking emits bounded deterministic markup without changing report data", () => {
+  // Given: reviewed phrases plus escaping, long-token, path, URL, and mixed-Latin boundaries.
+  const fixture = sampleFindings();
+  const reportData = {
+    ...fixture,
+    language: "zh-CN",
+    target: { name: "render-fixture", path: "/tmp/render-fixture" },
+    summary: {
+      ...fixture.summary,
+      overview: "同一证据包给出两套互相冲突的口径，并保留 recommendedReads 与 app/api/example.py。",
+      dimensions: fixture.summary.dimensions.map((dimension, index) => ({
+        ...dimension,
+        label: ["任务理解", "可控执行", "改动验证", "可靠交付", "经验沉淀"][index],
+        summary: index === 1
+          ? "项目具有工作入口，但画像没有投影这些命令，代理仍需猜测验证路线。"
+          : dimension.summary,
+      })),
+    },
+    findings: fixture.findings.map((finding, index) => ({
+      ...finding,
+      title: index === 0 ? "同一证据包给出相反结论" : "项目画像漏掉工作入口",
+      reason: index === 0
+        ? "安全<script>与命令：https://example.com/复核/very-long-token"
+        : `${finding.reason} 超长连续汉字测试文本边界仍可读取`,
+    })),
+  };
+
+  // When: the Chinese and English documents are rendered through the same HTML renderer.
+  const chineseHtml = renderHtml(reportData);
+  const englishHtml = renderHtml({ ...reportData, language: "en" });
+
+  // Then: ordinary Chinese word segments receive bounded keep-together markup.
+  assert.match(chineseHtml, /<html lang="zh-CN" class="no-js">/u);
+  for (const phrase of ["口径", "入口", "结论", "命令"]) {
+    assert.match(chineseHtml, new RegExp(`<span class="cjk-phrase">${phrase}</span>`, "u"));
+  }
+  const markedPhrases = [...chineseHtml.matchAll(/<span class="cjk-phrase">([^<]+)<\/span>/gu)];
+  assert.ok(markedPhrases.length > 4);
+  for (const [, phrase] of markedPhrases) {
+    assert.match(phrase, /^\p{Script=Han}{2,8}$/u);
+    assert.ok([...phrase].length <= 8);
+  }
+  const metricMarkup = chineseHtml.match(/<div class="metric">([\s\S]*?)<\/div>/u)?.[1] ?? "";
+  const scoreOrbitMarkup = chineseHtml.match(/<div class="score-orbit">([\s\S]*?)<\/div>/u)?.[1] ?? "";
+  const evidenceMarkup = chineseHtml.match(/<div class="evidence-grid">([\s\S]*?)<\/div>/u)?.[1] ?? "";
+  for (const surfaceMarkup of [metricMarkup, scoreOrbitMarkup, evidenceMarkup]) {
+    assert.match(surfaceMarkup, /<span>[\s\S]*?<span class="cjk-phrase">/u);
+  }
+  assert.match(chineseHtml, /\.score-orbit > span\s*\{[^}]*width:110px/u);
+  assert.match(chineseHtml, /\.metric > span,\.metric > small\s*\{[^}]*display:block/u);
+  assert.match(chineseHtml, /\.metric > strong\s*\{[^}]*display:block/u);
+  assert.match(chineseHtml, /\.evidence-grid > div > span,\.evidence-grid > div > strong\s*\{[^}]*display:block/u);
+  assert.match(chineseHtml, /\.evidence-grid > div > span\s*\{[^}]*font-size:12px/u);
+  assert.match(chineseHtml, /\.evidence-grid > div > strong\s*\{[^}]*margin-top:4px/u);
+  assert.match(chineseHtml, /<span class="cjk-phrase">安全<\/span>&lt;script&gt;/u);
+  assert.doesNotMatch(chineseHtml, /安全<script>/u);
+  assert.match(chineseHtml, /recommendedReads/u);
+  assert.match(chineseHtml, /app\/api\/example\.py/u);
+  assert.match(chineseHtml, /https:\/\/example\.com\//u);
+
+  // And: English output and nonvisual report surfaces do not gain phrase markup.
+  assert.match(englishHtml, /<html lang="en" class="no-js">/u);
+  assert.doesNotMatch(englishHtml, /<span class="cjk-phrase">/u);
+  assert.doesNotMatch(chineseHtml, /aria-label="[^"]*<span class="cjk-phrase">/u);
+  assert.doesNotMatch(chineseHtml, /<script id="harness-report-data" type="application\/json">[\s\S]*?<span class="cjk-phrase">/u);
+
+  // And: unsupported runtimes fall back to escaped readable text.
+  const segmenter = Intl.Segmenter;
+  let fallbackHtml;
+  try {
+    Intl.Segmenter = undefined;
+    fallbackHtml = renderHtml(reportData);
+  } finally {
+    Intl.Segmenter = segmenter;
+  }
+  assert.doesNotMatch(fallbackHtml, /<span class="cjk-phrase">/u);
+  assert.match(fallbackHtml, /安全&lt;script&gt;与命令/u);
+});
+
 test("HTML validator rejects incomplete finding action contracts", () => {
   const reportData = {
     ...sampleFindings(),
diff --git a/test/session-population.test.mjs b/test/session-population.test.mjs
new file mode 100644
index 0000000..1f30f78
--- /dev/null
+++ b/test/session-population.test.mjs
@@ -0,0 +1,90 @@
+import assert from "node:assert/strict";
+import test from "node:test";
+
+async function populationModule() {
+  return import("../scripts/session-analysis/session-population.mjs");
+}
+
+test("frozen population binding omits exact active and Qoder home-only sessions", async () => {
+  const { freezeSessionPopulation } = await populationModule();
+  const population = freezeSessionPopulation({
+    scope: {
+      platform: "qoder",
+      workspace: "/private/workspace",
+      since: "2026-07-01T00:00:00.000Z",
+      until: "2026-07-30T00:00:00.000Z",
+    },
+    sessions: [
+      { sessionId: "active-private", sourceRefs: [{ kind: "project-session", path: "/private/active.jsonl" }] },
+      { sessionId: "home-private", sourceRefs: [{ kind: "home-session", path: "/private/home.jsonl" }] },
+      { sessionId: "eligible-private", sourceRefs: [{ kind: "project-session", path: "/private/eligible.jsonl" }] },
+    ],
+    excludedSessionId: "active-private",
+    suppliedUntil: true,
+  });
+
+  assert.deepEqual(population.sessions.map((session) => session.sessionId), ["eligible-private"]);
+  assert.equal(population.binding.omission.activeSessions, 1);
+  assert.equal(population.binding.omission.homeSessionOnly, 1);
+  assert.equal(population.binding.omission.exactIdentityAvailable, true);
+  assert.equal(population.binding.omission.recencyInference, "disabled-frozen-until");
+  assert.equal(population.binding.eligible.count, 1);
+  assert.match(population.binding.eligible.fingerprint, /^[a-f0-9]{16}$/u);
+  assert.doesNotMatch(JSON.stringify(population.binding), /active-private|home-private|eligible-private|\/private/u);
+});
+
+test("selection binding rejects a session outside the frozen population", async () => {
+  const { bindSessionSelection, freezeSessionPopulation } = await populationModule();
+  const population = freezeSessionPopulation({
+    scope: { platform: "codex", workspace: "/workspace", until: "2026-07-30T00:00:00.000Z" },
+    sessions: [{ sessionId: "eligible" }],
+    suppliedUntil: true,
+  });
+
+  assert.throws(
+    () => bindSessionSelection(population, [{ sessionId: "foreign" }], {
+      strategy: "stratified",
+      projectionPolicy: "lead-report-signal-v1",
+    }),
+    (error) => error?.code === "SESSION_SELECTION_OUTSIDE_POPULATION",
+  );
+});
+
+test("binding validation preserves zero-signal lead admission and reconciles Session facts", async () => {
+  const {
+    bindSessionSelection,
+    freezeSessionPopulation,
+    sessionAdmissionBinding,
+    leadAdmissionBinding,
+    validateSessionPopulationBundle,
+  } = await populationModule();
+  const population = freezeSessionPopulation({
+    scope: { platform: "codex", workspace: "/workspace", until: "2026-07-30T00:00:00.000Z" },
+    sessions: [{ sessionId: "eligible" }],
+    suppliedUntil: true,
+  });
+  const sessionSelection = bindSessionSelection(population, population.sessions, {
+    strategy: "all-eligible",
+    projectionPolicy: "session-fact-candidates-v2",
+  });
+  const leadSelection = bindSessionSelection(population, population.sessions, {
+    strategy: "stratified",
+    projectionPolicy: "lead-report-signal-v1",
+  });
+  const sessionAdmission = sessionAdmissionBinding({
+    admission: { taskEpisodes: 1, candidateEpisodes: 1, distinctRequests: 1, emittedCandidates: 1 },
+    omitted: { noRequest: 0, selfAnalysis: 0, lowSignal: 0, duplicateRequests: 0, candidateBudget: 0 },
+  }, sessionSelection);
+  const leadAdmission = leadAdmissionBinding({
+    projectedEpisodes: 1,
+    admittedEpisodes: 0,
+    zeroSignalDiscardedEpisodes: 1,
+    retainedTaskEpisodes: 0,
+  }, leadSelection);
+
+  assert.deepEqual(validateSessionPopulationBundle({
+    population: population.binding,
+    session: { population: population.binding, selection: sessionSelection, admission: sessionAdmission },
+    lead: { population: population.binding, selection: leadSelection, admission: leadAdmission },
+  }), []);
+});
diff --git a/test/task-loop-source.test.mjs b/test/task-loop-source.test.mjs
index 679d1a0..cb66352 100644
--- a/test/task-loop-source.test.mjs
+++ b/test/task-loop-source.test.mjs
@@ -5,6 +5,7 @@ import path from "node:path";
 import test from "node:test";
 
 import { buildCheckupScan } from "../scripts/coding-agent-practices/checkup/scan.mjs";
+import { freezeSessionPopulation } from "../scripts/session-analysis/session-population.mjs";
 import {
   buildHarnessReviewPacket,
   validateHarnessReviewPacket,
@@ -396,7 +397,7 @@ test("source generation keeps usage opt-in and rejects incomplete requested cens
   );
 });
 
-test("requested usage reuses one cutoff and the initially discovered session inventory", async () => {
+test("requested usage reuses one frozen population without rediscovery and emits its lead selection binding", async () => {
   const root = await mkdtemp(path.join(os.tmpdir(), "better-harness-frozen-usage-"));
   const workspace = path.join(root, "workspace");
   const calls = [];
@@ -405,7 +406,7 @@ test("requested usage reuses one cutoff and the initially discovered session inv
     firstSeen: "2026-07-17T08:00:00.000Z",
     lastSeen: "2026-07-17T08:05:00.000Z",
     sourceKinds: ["fixture"],
-    sourceRefs: [],
+    sourceRefs: [{ kind: "project-session", path: "/fixture/session.jsonl" }],
   }));
   const activity = {
     schemaVersion: 1,
@@ -489,12 +490,22 @@ test("requested usage reuses one cutoff and the initially discovered session inv
   };
   try {
     await mkdir(workspace, { recursive: true });
-    const { source, selection } = await createTaskLoopSourceFromSessions({
+    const sessionPopulation = freezeSessionPopulation({
+      scope: {
+        platform: "qoder",
+        workspace,
+        until: "2026-07-17T09:00:00.000Z",
+      },
+      sessions: initialSessions,
+      suppliedUntil: true,
+    });
+    const { source, selection, sessionBinding } = await createTaskLoopSourceFromSessions({
       analyzer,
       platform: "qoder",
       workspace,
       snapshotUntil: "2026-07-17T09:00:00.000Z",
       includeUsage: true,
+      sessionPopulation,
       qoderHome: path.join(root, ".qoder"),
       piHome: path.join(root, ".pi", "agent"),
       practiceInventory: { summary: { practiceCoverageRows: [] }, memories: { included: false, categories: [] } },
@@ -502,8 +513,13 @@ test("requested usage reuses one cutoff and the initially discovered session inv
     assert.equal(selection.eligibleCount, 2);
     assert.equal(source.sessionEvents.usageActivity.sessions.total, 2);
     assert.equal(source.sessionEvents.usageEfficiency.selection.eligibleSessionCount, 2);
+    assert.equal(sessionBinding.population.eligible.count, 2);
+    assert.equal(sessionBinding.selection.parentPopulationFingerprint, sessionPopulation.binding.eligible.fingerprint);
+    assert.equal(sessionBinding.admission.projectedEpisodes, sessionBinding.admission.admittedEpisodes
+      + sessionBinding.admission.zeroSignalDiscardedEpisodes);
     assert.equal(new Set(calls.map((call) => call.until)).size, 1);
     assert.ok(calls.every((call) => call.piHome === path.join(root, ".pi", "agent")));
+    assert.equal(calls.filter((call) => call.command === "sources").length, 0);
     assert.deepEqual(calls.filter((call) => call.command === "insights").map((call) => call.inventory), [
       ["session-a", "session-b"],
       ["session-a", "session-b"],