Repository navigation
Reconcile PersonalOS instructions with the six-plugin suite - #77
Conversation
|
🚀 personal-ai-chat build ready for testing Artifact: Download from workflow run Playground ZIP: Launch in Playground: Open WordPress Playground Paste the ZIP URL into a WordPress Playground blueprint to load this PR build. |
|
🚀 personal-notes build ready for testing Artifact: Download from workflow run Playground ZIP: Launch in Playground: Open WordPress Playground Paste the ZIP URL into a WordPress Playground blueprint to load this PR build. |
|
🚀 personal-readwise-sync build ready for testing Artifact: Download from workflow run Playground ZIP: Launch in Playground: Open WordPress Playground Paste the ZIP URL into a WordPress Playground blueprint to load this PR build. |
|
🚀 personal-evernote-sync build ready for testing Artifact: Download from workflow run Playground ZIP: Launch in Playground: Open WordPress Playground Paste the ZIP URL into a WordPress Playground blueprint to load this PR build. |
|
🚀 personal-todo build ready for testing Artifact: Download from workflow run Playground ZIP: Launch in Playground: Open WordPress Playground Paste the ZIP URL into a WordPress Playground blueprint to load this PR build. |
|
🚀 personal-stuff build ready for testing Artifact: Download from workflow run Playground ZIP: Launch in Playground: Open WordPress Playground Paste the ZIP URL into a WordPress Playground blueprint to load this PR build. |
The root instructions mixed legacy monolith setup with the current six-plugin suite and accumulated contradictory lessons. This makes contributors choose between incompatible rules, including when Stuff creates a row and where structural taxonomy terms may be assigned.
Replace the append-only guide with a short task router and scoped Knowledge, development, packaging, sync, Notes and AI Chat contracts. Preserve existing Stuff and TODO contracts and their UI guidance, refresh the README for standalone packages, and retire Cursor-specific rules in favor of the portable
AGENTS.mdentrypoint.CLAUDE.mdand the shared skills compatibility link continue to expose the same guidance to Claude.Routine authorized local environment repair is allowed, with deployment still separately scoped. This changes documentation only. Notes creation and AI Chat response-shape failures remain verification requirements; this PR does not fix or claim to fix runtime bugs.
Validation: checked all 29 local Markdown links, verified referenced npm scripts and package-selection flags against source, verified the
AGENTS.md/Claude/skills discovery links, removed stale Cursor-rule references, and passedgit diff --check. No application tests were run for this documentation-only change.