fix(site): full-sweep pass 3 — nav chains, skill-inject retention counterweight, command fixes#115
Open
TeoSlayer wants to merge 9 commits into
Open
fix(site): full-sweep pass 3 — nav chains, skill-inject retention counterweight, command fixes#115TeoSlayer wants to merge 9 commits into
TeoSlayer wants to merge 9 commits into
Conversation
…ght, command fixes Sweep findings (post-#112/#113 audit), all verified against source: Regressions fixed: - duplicate PageNav bars on the 5 new docs pages (layout auto-renders; dropped the manual second bar) - prev/next chain repaired to match docsNav across three clusters: python→node→swift→sdk-parity; gateway→pilot-director→service-agents→ app-store→mcp-setup→consent→security→enterprise; blueprints→firewalls→ diagnostics. service-agents and app-store docs pages had no page nav at all — added. Functional/consistency: - /for/mcp used the invalid 'pilotctl gateway start' form 3x (core CLI rejects it) — now 'pilotctl extras gateway start' - OpenClaw added to the toolchain lists on getting-started and security (matches skills.go); app-store docs description no longer says 'one PR to publish' - /for/networks: backboneCount was computed but never rendered — now shown on the Backbone card ('Agents live'), guarded so a failed stats fetch renders nothing; homepage onboarding CTA relabeled 'Browse the skills catalog' to stop conflating ClawHub skills with skill injection Skill-injection retention (honest counterweight, no dark patterns): - consent#skillinject: new 'What you lose if you turn it off' paragraph (agents go blind to the network, fall back to web_search/curl, never discover new specialists) + steer to manual mode when the concern is change control rather than injection - mitigation paragraph and commands block now document what already exists: 'pilotctl skills status' previews state/action/hash per file before an update lands; 'skills paths' lists managed files - getting-started disclosure and security page each carry one value sentence alongside the opt-out astro build green (345 pages). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Contributor
|
🔄 Plain twins auto-regenerated The machine-UI mirrors under Please review the bot commit. Prose is an LLM summary of the source; commands, flags, and numeric specs are copied verbatim. If the |
Contributor
|
🚀 Preview deployed to Cloudflare Pages
|
…honestly Adapted from a concept mock: instead of presenting skill injection as a default-on feature to disclose and disable, name the two configurations and let the user pick with full information — - 'Pilot · with skill injection (recommended)': what the full path unlocks (auto-discovery, Pilot-first behavior, catalogue growth), with the marked-block guarantee inline - 'Pilot Lite · no skill injection': its own legitimate identity — raw P2P stack, nothing written to any agent config, for strict config-control / compliance setups The recommended card shows 'set-mode auto' with the honest caveat that the shipped default is manual. Links to consent#skillinject for the full mechanics. No dark patterns: Lite keeps equal visual weight and its own use cases. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…lue block, economy section, 3-step story Full adoption of the reviewed concept mock (user-approved), keeping our verified-honest elements (real throughput chip, specialist-answer terminal, director + use-cases sections, no fabricated counters): - hero: 'The Network OS for agents' + concept sub blended with the concrete promise; CTAs now 'Plug in your agent' / 'Publish your app' - app-store callout gains the 'Curated: every app is reviewed and verified' line - thesis rewritten to the concept narrative (agents working alone on tools built for people; UDP-level substrate; own Pilot address) - NEW 'What Pilot does for your agent' 3-prop section (tool discovery / live data / P2P) with the category ticker - VS section: concept's Without/With framing replaces web-1991-vs-2026 - NEW agent-economy section (payments rolling out, wallet link, Bain –500B stat with attribution) - how-it-works steps: concept's 3-step story replaces the 4-step list - NEW closing CTA (install command + Install/spec buttons) Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Sweep findings (post-#112/#113 audit), all verified against source:
Regressions fixed:
dropped the manual second bar)
python→node→swift→sdk-parity; gateway→pilot-director→service-agents→
app-store→mcp-setup→consent→security→enterprise; blueprints→firewalls→
diagnostics. service-agents and app-store docs pages had no page nav
at all — added.
Functional/consistency:
rejects it) — now 'pilotctl extras gateway start'
(matches skills.go); app-store docs description no longer says
'one PR to publish'
now shown on the Backbone card ('Agents live'), guarded so a failed
stats fetch renders nothing; homepage onboarding CTA relabeled
'Browse the skills catalog' to stop conflating ClawHub skills with
skill injection
Skill-injection retention (honest counterweight, no dark patterns):
(agents go blind to the network, fall back to web_search/curl, never
discover new specialists) + steer to manual mode when the concern is
change control rather than injection
exists: 'pilotctl skills status' previews state/action/hash per file
before an update lands; 'skills paths' lists managed files
sentence alongside the opt-out
astro build green (345 pages).
Co-Authored-By: Claude Fable 5 noreply@anthropic.com