Skip to content

Releases: NickCollect/ai-coding-runbook

Release list

v0.1.0 — Initial public release

v0.1.0 — Initial public release Latest
Latest

Choose a tag to compare

View all tags
@NickCollect NickCollect released this 05 May 19:31
v0.1.0
75dc258

First public release.

Numbers below are accurate as of v0.1.0 (2026-05-05). main will drift over time as new content lands; see README for current stats.

What's in the box

  • 01_Raw/ — 11,500+ raw files (1,650+ markdown + git-cloned source code) auto-synced from 9 GHA matrix sources: Claude Code docs, Anthropic API + platform docs, anthropic.com (research/engineering blog), Cursor IDE docs (cursor.com/docs), Gemini API docs, openai.com blog, plus shallow-cloned GitHub repos from anthropics/*, modelcontextprotocol/*, and openai/*. Plus 30 manually fetched key pages from platform.openai.com/docs (Cloudflare-blocked from auto-crawling).
  • 02_Wiki/ — 1,358 LLM-enriched summaries, 85 entity dossiers, 25 concept files, 8 synthesis essays, 5 cross-vendor decision matrices, 7 Q&A docs.
  • 03_Output/ — 10 hand-maintained cheatsheets (hook recipes, slash command refs, plugin install guides, Skill vs Plugin vs MCP vs Subagent decision matrix, etc.) and auto-generated weekly changelogs.

Automation

  • GitHub Actions matrix-parallel cron, every Monday 09:00 HKT (~10 min wall clock; first verified end-to-end run 2026-05-05, 9/9 success, 12m42s).
  • Per-source independent commits with git pull --rebase retry-loop for safe concurrent push.
  • Auto-generated changelog after each refresh aggregates all bot commits in the past 2h.

How to use

  1. git clone https://github.com/NickCollect/ai-coding-runbook && cd ai-coding-runbook
  2. Open the folder in your AI agent of choice (Claude Code / Cursor / Codex CLI / Gemini CLI / Aider).
  3. Ask: "What's the difference between Skills, MCP servers, and Subagents?"

The agent auto-loads CLAUDE.md / AGENTS.md and answers from the enriched wiki. Zero config, zero API key (the agent uses your own subscription).

Documentation

  • README.md / README.en.md — landing page (Why / Quickstart / Status / Limitations).
  • docs/ARCHITECTURE.md — full directory tree + 5 core mechanisms.
  • docs/INGEST_WORKFLOW.md — LLM ingest SOP (Phase A→E + daily user workflow).
  • docs/MAINTENANCE.md — maintainer's manual (lessons learned, ops cadence, glossary).

License

MIT. 01_Raw/ is a third-party content cache — copyright belongs to original authors (Anthropic / OpenAI / Google / Cursor / etc.); this project mirrors them locally + produces derivative summaries.

Known limitations

  • Weekly refresh; not real-time.
  • Doesn't pay for LLM APIs — all enrichment is user-triggered in your own agent session.
  • platform.openai.com/docs is Cloudflare-blocked; OpenAI is covered via GitHub repos + 30 manually fetched key pages.
  • v0.1 — small contributor base, limited external user feedback so far.

See README for full details.

Assets 2
Loading