Skip to module content
🧠AI Academy
0/91
Sign in
Module 24 Β· ~5 min

Master Resource Index

Every repo, directory, doc, and person worth bookmarking.

Reading progress
0/8 Β· 0%

The big idea

πŸ’‘Key idea
Access to AI resources is no longer scarce. Curation is. The index organizes what's worth bookmarking across six categories: Official Anthropic (anthropics/skills, claude-code-action, Plugin Directory, engineering posts); the skills ecosystem (awesome-claude-code, ComposioHQ, travisvn, superpowers, agents-best-practices, Ay-Skills, Skill_Seekers); MCP directories and tooling (Smithery, Glama, PulseMCP, FastMCP, MetaMCP, local-mcp-gateway); GTM/lead-gen (clay.com/blog, Clay outreach bot, Amplemarket); research tools (Firecrawl, Exa, Tavily, Perplexity Sonar, Parallel, Linkup, Brave); and GEO and context/evals resources. The rule for every item: one-line adoption plan or it doesn't make the list. Bookmarks without intent are noise.
Quick check
1 question Β· instant feedback
0/1
  1. The de-facto index of the ecosystem to review monthly is:

Numbers that matter

6
Bookmark categories: Anthropic official, skills ecosystem, MCP, GTM/lead-gen, research, GEO.
Quick check
1 question Β· instant feedback
0/1
  1. Own assets should be:

Deep dive

3/3 open

Start here before building anything from scratch.

**anthropics/skills** β€” Anthropic's own reference implementations. Read these before writing your own skills; they set the standard for structure, safety, and progressive disclosure.

**claude-code-action** β€” the headless/CI pattern for running Claude in automated pipelines.

**Official Plugin Directory** β€” the canonical list of what's been built and sanctioned.

**Engineering posts worth reading closely:** 'How Claude Code Works' (architecture), 'How We Built Our Multi-Agent Research System' (canonical orchestrator/subagent design β€” this is the reference for Sprint 6), 'Steering Claude Code' (the extension decision framework: when to use hooks vs skills vs subagents), and 'Skills Explained' (progressive disclosure + decision matrices).

**docs.claude.com** β€” the live reference for Agent SDK, Skills API, MCP docs, and the prompt engineering guide. Check it before trusting anything older than three months.

These are the repos and resources that the builder community actually uses β€” a better signal than vendor marketing.

**hesreallyhim/awesome-claude-code** β€” the de-facto index. Review it monthly. It's the fastest way to see what's gaining traction in the ecosystem.

**ComposioHQ/awesome-claude-skills + travisvn/awesome-claude-skills** β€” two curated skill libraries worth cross-referencing.

**obra/superpowers** (~94k stars) and **'Everything Claude Code'** (~100k-star class) β€” high-signal aggregators.

**awesome-claude-code-toolkit** β€” hooks, commands, plugins; the practical layer.

**DenisSergeevitch/agents-best-practices** β€” the harness engineering reference. This is where Module 18's risk taxonomy language comes from.

**Ay-Skills** (SEO suite), **yusufkaraaslan/Skill_Seekers** (docs→skill converter), **glebis/claude-skills** (retrospective and consulting context-builder).

Also worth studying: **x1xhlol/system-prompts-and-models-of-ai-tools** β€” a reverse-engineered collection of how the best AI products actually prompt. Study these; don't copy them.

The most overlooked item in the index is what you've already built.

Inventory your own assets the same way you'd catalog an external repo β€” with a name, a description, and an adoption/offer plan.

What to include: the eight-skill Google Ads suite (pull/push/retract) — this is the Sprint 3 reference implementation; the custom LinkedIn MCP; the Meta+Google+Canva+Gmail+Drive+Apollo+Porter stack; the gateway evaluation; the DTI digital-twin platform (Supabase edge functions + Claude API + Lovable, no middleware) — this is the Digital Twin offer's proof of concept; the LinkedIn→Canva carousel pipeline plus GMS Templates/Posts structure — the seed of every Content OS sprint; the Lovable-built outreach CRM — the internal-tools pattern.

These aren't side projects. They're the raw material for every sprint you sell. Treat them accordingly.

Quick check
1 question Β· instant feedback
0/1
  1. Bookmarks without an adoption plan are:

How to run it

  1. Monthly review of awesome-claude-code
    Fastest way to see what builders actually use.
  2. One-line adoption plan per bookmark
    No bookmarks without intent.
  3. Inventory own assets as IP
    They're the raw material for offers.
  4. Fork every adopted repo
    Vendor + pin.
Quick check
1 question Β· instant feedback
0/1
  1. The closing thesis of the manual is:

In the field

πŸ”¬Worked example
Your own assets as IP: eight-skill Google Ads suite, custom LinkedIn MCP, Meta+Google+Canva+Gmail+Drive+Apollo+Porter stack, DTI digital-twin platform, LinkedIn→Canva carousel pipeline, Lovable outreach CRM. Inventory your assets — they're the raw material for reusable offers.
🚫When not to reach for it
Don't add a resource to the bench without an intended use. Bookmarks without adoption plans become noise.
Quick check
1 question Β· instant feedback
0/1
  1. The 'Steering Claude Code' post is important because it:

Pitfalls & takeaways

Failure modes

  • Bookmark hoarding without adoption plans.
  • Not inventorying your own assets as IP.
  • Missing the monthly review of hesreallyhim/awesome-claude-code.
  • Failing to update the index as the ecosystem shifts.

Durable takeaways

  • Curation is the scarce resource β€” not access.
  • Review awesome-claude-code monthly; the ecosystem moves fast.
  • Your own assets are IP; inventory them like external resources.
  • Bookmarks without adoption plans are noise.
  • The market moved from 'AI tools' to 'AI operating systems' β€” GMS named the offer correctly before the market caught up.

Do the work

πŸ‹οΈProve you learned it

Fork/bookmark the top 10 resources from this index and add one-line adoption plans for each. Inventory your own assets in the same format. Set a monthly calendar reminder to review awesome-claude-code and update your index.

0 chars
πŸ“¦Artifact to produce
Personal resource index with monthly review cadence + own-asset inventory.

Sources

  • Β· GMS Field Manual Β§24 (Master Resource Index)