Conversation
NotebookLM scouts: SEO farm detection, broad-to-narrow search strategy NotebookLM workers: SEO-suspect handling, parallel notebook_query NotebookLM command: delegation clarity checklist Research docs: claims verification, plan review, implementation tracking Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Fixed unreachable case patterns in context-pressure-advisory.sh (SC2221/SC2222) - Suppressed intentional glob expansion warning in find-polluter.sh (SC2206) - Routed codex-review-gate through /codex:rescue instead of /codex:review (disable-model-invocation) - Updated workday-complete and bug-sweep to match Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…kill, all refs cleaned, v1.3.0
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 564d86f497
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| @@ -0,0 +1,28 @@ | |||
| --- | |||
| description: "Run a deep research pipeline on a topic across internet sources (Pipeline A), a repository (Pipeline B), or structured research with schema-conforming output (Pipeline C). Use for studying codebases, building knowledge bases, evaluating libraries, or investigating multi-source technical topics with verified findings. For batch structured research campaigns, use /structured-research instead." | |||
| allowed-tools: ["Read", "Bash"] | |||
There was a problem hiding this comment.
This router’s Step 2 requires the Skill tool to dispatch deep-research:web, deep-research:repo, or deep-research:structured, but the front matter only permits Read and Bash. In practice that blocks the dispatch call, so /deep-research ... can’t hand off to any pipeline and fails at routing time.
Useful? React with 👍 / 👎.
| "version": "1.1.0", | ||
| "description": "Turn Claude Code into a structured development partner — session lifecycle, named code reviewers, research pipelines, and 24 workflow skills that enforce quality from plan through execution.", | ||
| "license": "MIT", | ||
| "version": "1.0.0", |
There was a problem hiding this comment.
The manifest version was reduced to 1.0.0 even though this commit introduces new coordinator changes. Downgrading the version can prevent existing installs on newer cached versions from recognizing this as an update, which leaves users on stale command/skill behavior. The plugin version should continue increasing, not regress.
Useful? React with 👍 / 👎.
1 participant
564d86f pre-merge quick-save
03696fb D-032: superpowers conscious uncoupling — independent brainstorming skill, all refs cleaned, v1.3.0
1189a00 deep-research: Pipeline B survey, atlas sketch, deduplication (synced from ~/.claude)
2ae3f84 remove notebooklm (moved to deep-research), update pipeline refs
2be5928 back-port: notebooklm research.md from installed
1be01cb fix: README inventory counts, remember skill YAML list, /daily-review entry
dfd592e sync from ~/.claude: shellcheck fixes, codex review gate routing
30a6e50 notebooklm + research docs: apply Anthropic multi-agent lessons