Problem
Bigpowers generates rich YAML specs (epics, stories, tasks, metrics, bugs) but downstream Allure progress dashboards only consume ~30% of the available data. The gap is in how skills write YAML outputs — key metadata (risk, security, WSJF, cycle times, task-level status) exists in the bigpowers schema but is either not consistently written or not structured for downstream consumption.
Current Data Flow
scope-work → slice-tasks → plan-work → build-epic → develop-tdd → release-branch
↓ ↓ ↓ ↓ ↓ ↓
SCOPE_ epic.yaml eNNsYY- execution- tasks.yaml cycle-times.yaml
LATEST.yaml stories[] tasks.yaml status.yaml status: + OKF metrics
risk: P0-P3 failing→passing
What Bigpowers Has But Dashboards Cannot See
| Data |
Location |
Why Missing |
risk: P0/P1/P2/P3 |
tasks[].risk |
Generator ignores it |
security: none/low/medium/high |
tasks[].security |
Generator ignores it |
wsjf |
release-plan.yaml |
Not read by generator |
tier |
release-plan.yaml |
Not read by generator |
wave |
release-plan.yaml |
Not read by generator |
depends_on |
epic.yaml stories |
Not read by generator |
delta: ADDED/MODIFIED/REMOVED |
epic.yaml stories |
Not read by generator |
total_bcps |
epic.yaml |
Not read by generator |
start/end timestamps |
cycle-times.yaml |
Not read by generator |
effort_hours |
OKF bundle |
Not read by generator |
| DORA metrics |
OKF bundle |
Not read by generator |
| Bug registry |
specs/bugs/registry.yaml |
Not read by generator |
status: failing/passing per task |
tasks[].status |
Only story-level status checked |
Root Cause: Three Disconnects
1. Task-Level Status Not Surfaced
plan-work writes status: failing on each task. develop-tdd flips to status: passing. But sync-status-from-epics.sh only reads story-level status from epic.yaml — it never reads per-task status from eNNsYY-tasks.yaml. Dashboards see all-or-nothing at the story level.
2. Rich Metadata Written But Not Indexed
plan-release writes wsjf, tier, wave, depends_on to release-plan.yaml. plan-work writes risk, security to task YAML. But sync-status-from-epics.sh produces a flat key: status dict — all metadata is lost in aggregation.
3. Cycle Times Siloed
record-cycle-time.sh produces rich OKF bundles with DORA metrics, agent telemetry, and effort data in specs/metrics/cycle-times.yaml — completely separate from the execution-status pipeline.
Proposed Changes
A. sync-status-from-epics.sh — Emit Richer Metadata
Current output:
epics:
e01: done
stories:
e01s01: done
Proposed output:
epics:
e01:
status: done
title: "CLI — BigBase Breathes"
wsjf: 1.0
tier: 1
total_bcps: 5
stories:
e01s01:
status: done
title: "Login"
bcps: 3
epic: e01
tasks_total: 4
tasks_passing: 4
tasks_failing: 0
risk_max: P1
security_max: none
B. plan-work — Write allure_categories in Task YAML
tasks:
- id: 1
description: "Add login form"
verify: "npm test -- login.test.tsx"
risk: P1
security: none
status: failing
allure:
severity: critical # P0→critical, P1→high, P2→normal, P3→minor
categories:
- "Security Review"
- "Wave 1"
C. build-epic — Record Story Start/End in Execution Status
On story complete, also write timestamps:
stories:
e01s01:
status: done
started_at: "2026-07-12T16:00:00-03:00"
completed_at: "2026-07-12T18:45:00-03:00"
D. record-cycle-time.sh — Write to Execution Status
Also update execution-status.yaml with cycle time data:
stories:
e01s01:
status: done
effort_hours: 1.5
lead_time_minutes: 90
bcp_per_hour: 2.0
E. plan-release — Write Bug Registry Summary
Add to release-plan.yaml:
bugs:
total: 42
fixed: 38
deferred: 2
wontfix: 2
registry: specs/bugs/registry.yaml
F. New Skill: generate-allure-report (Bigpowers Native)
Ship a built-in skill that:
- Reads
execution-status.yaml, release-plan.yaml, epic capsules, task YAMLs, cycle-times.yaml, and bugs/registry.yaml
- Produces
allure-results/junit-results.xml with <properties> for risk, security, WSJF, tier, wave
- Produces
allure-results/categories.json for custom Allure categories
- Produces
allure-results/executor.json with build metadata from state.yaml
What Already Works (No Changes Needed)
- Epic/Story/Task hierarchy in
epic.yaml
- WSJF scoring in
release-plan.yaml
- Risk classification (
tasks[].risk: P0-P3)
- Security annotations (
tasks[].security)
- Dependency graph (
depends_on, blocks)
- Delta tracking (
delta: ADDED/MODIFIED/REMOVED)
- Cycle time metrics in
specs/metrics/cycle-times.yaml
- Bug registry in
specs/bugs/registry.yaml
- Task status lifecycle (
failing → passing)
Impact
- All proposed additions are backward-compatible — existing consumers of
execution-status.yaml continue to work with flat key: status format
- No schema breaking changes — richer structure is additive
- Eliminates per-project Allure glue code — projects stop writing their own
generate-allure-report.py
Suggested Milestone
v2.80.0 or next minor release
Problem
Bigpowers generates rich YAML specs (epics, stories, tasks, metrics, bugs) but downstream Allure progress dashboards only consume ~30% of the available data. The gap is in how skills write YAML outputs — key metadata (risk, security, WSJF, cycle times, task-level status) exists in the bigpowers schema but is either not consistently written or not structured for downstream consumption.
Current Data Flow
What Bigpowers Has But Dashboards Cannot See
risk: P0/P1/P2/P3tasks[].risksecurity: none/low/medium/hightasks[].securitywsjfrelease-plan.yamltierrelease-plan.yamlwaverelease-plan.yamldepends_onepic.yamlstoriesdelta: ADDED/MODIFIED/REMOVEDepic.yamlstoriestotal_bcpsepic.yamlstart/endtimestampscycle-times.yamleffort_hoursspecs/bugs/registry.yamlstatus: failing/passingper tasktasks[].statusRoot Cause: Three Disconnects
1. Task-Level Status Not Surfaced
plan-workwritesstatus: failingon each task.develop-tddflips tostatus: passing. Butsync-status-from-epics.shonly reads story-level status fromepic.yaml— it never reads per-task status fromeNNsYY-tasks.yaml. Dashboards see all-or-nothing at the story level.2. Rich Metadata Written But Not Indexed
plan-releasewriteswsjf,tier,wave,depends_ontorelease-plan.yaml.plan-workwritesrisk,securityto task YAML. Butsync-status-from-epics.shproduces a flatkey: statusdict — all metadata is lost in aggregation.3. Cycle Times Siloed
record-cycle-time.shproduces rich OKF bundles with DORA metrics, agent telemetry, and effort data inspecs/metrics/cycle-times.yaml— completely separate from the execution-status pipeline.Proposed Changes
A.
sync-status-from-epics.sh— Emit Richer MetadataCurrent output:
Proposed output:
B.
plan-work— Writeallure_categoriesin Task YAMLC.
build-epic— Record Story Start/End in Execution StatusOn story complete, also write timestamps:
D.
record-cycle-time.sh— Write to Execution StatusAlso update
execution-status.yamlwith cycle time data:E.
plan-release— Write Bug Registry SummaryAdd to
release-plan.yaml:F. New Skill:
generate-allure-report(Bigpowers Native)Ship a built-in skill that:
execution-status.yaml,release-plan.yaml, epic capsules, task YAMLs,cycle-times.yaml, andbugs/registry.yamlallure-results/junit-results.xmlwith<properties>for risk, security, WSJF, tier, waveallure-results/categories.jsonfor custom Allure categoriesallure-results/executor.jsonwith build metadata fromstate.yamlWhat Already Works (No Changes Needed)
epic.yamlrelease-plan.yamltasks[].risk: P0-P3)tasks[].security)depends_on,blocks)delta: ADDED/MODIFIED/REMOVED)specs/metrics/cycle-times.yamlspecs/bugs/registry.yamlfailing→passing)Impact
execution-status.yamlcontinue to work with flatkey: statusformatgenerate-allure-report.pySuggested Milestone
v2.80.0 or next minor release