agentleFS
TakaGotoTakaGoto/rag-learning-academy/.claude/skills/audit-content/SKILL.md

audit-content

Skill24 starsChanged 19 days agoMITView on GitHub

Audit academy materials for outdated references, deprecated patterns, and stale content

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 1 is distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If the user specifies a scope (e.g., /audit-content embeddings or /audit-content module 3), audit only that area.

    line 12
  2. Very specific:

    Use WebSearch to check the MTEB leaderboard and model provider documentation for current recommendations.

    line 35
  3. Very specific:

    Use WebSearch to check changelogs and migration guides.

    line 50
  4. Distinctive:

    Check if major new RAG techniques have emerged that aren't covered

    line 61
  5. Very specific:

    Use WebSearch to check paper citation counts and newer follow-up work.

    line 63

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Audit Content: Keep the Academy Current

Purpose: Systematically check all academy materials for outdated information — deprecated models, changed APIs, superseded techniques, and stale references. Produces an actionable update report.

Step 1: Determine Audit Scope

If the user specifies a scope (e.g., /audit-content embeddings or /audit-content module 3), audit only that area. Otherwise, run a full audit across all content.

Full audit covers:

  • Curriculum modules (.claude/docs/curriculum/)
  • Agent definitions (.claude/agents/)
  • Reference docs (.claude/docs/reference/)
  • Code rules (.claude/rules/)
  • Sample data (data/raw/)
  • Templates (.claude/docs/templates/)

Step 2: Check Model References

Scan all content files for embedding and LLM model references. Verify each against current state:

Model ReferenceWhat to Check
text-embedding-3-smallStill recommended? Check OpenAI docs and MTEB leaderboard
all-MiniLM-L6-v2Still a good free default? Check Sentence Transformers releases
nomic-embed-textStill maintained and competitive?
bge-base-en-v1.5Check if newer BGE versions exist
Claude model IDsVerify against current Anthropic model IDs
GPT model IDsVerify against current OpenAI model IDs

Use WebSearch to check the MTEB leaderboard and model provider documentation for current recommendations.

Step 3: Check Library and Framework References

Scan for library references and verify versions and API patterns:

LibraryWhat to Check
langchainImport patterns — should use langchain-core, langchain-community, or langgraph
chromadbAPI compatibility — check for breaking changes
ragasCurrent version and metric names
sentence-transformersCurrent version and API
llama-indexPackage rename or API changes
pydanticv1 vs v2 patterns

Use WebSearch to check changelogs and migration guides.

Step 4: Check Paper and Technique References

Scan for research paper citations. Flag:

  • Papers described as "recent" or "new" that are 2+ years old
  • Techniques that have been superseded by newer approaches
  • Benchmarks that have been replaced

Key papers to verify:

  • Self-RAG, CRAG, HyDE, ColBERT, RAPTOR, GraphRAG
  • Check if major new RAG techniques have emerged that aren't covered

Use WebSearch to check paper citation counts and newer follow-up work.

Step 5: Check for Deprecated Patterns

Scan code examples and rules for deprecated patterns:

  • Old-style LangChain imports (from langchain import instead of from langchain_core)
  • Deprecated API parameters or function signatures
  • Outdated best practices

Step 6: Check Content Age

For each content file, check the git log to find when it was last updated. Flag files older than 90 days. Categorize by staleness risk:

Risk LevelReview CycleContent Types
HighMonthlyModel recommendations, library versions, MTEB references
MediumQuarterlyFramework patterns, API examples, production tooling
LowSemi-annuallyFoundational concepts, algorithms, evaluation theory

Step 7: Generate Audit Report

Produce a structured report:

# Content Audit Report — [Date]

## Summary
- Files audited: N
- Issues found: N (X critical, Y warnings, Z info)

## Critical (requires immediate update)
- [ ] Issue description — file path — suggested fix

## Warnings (review within 30 days)
- [ ] Issue description — file path — suggested fix

## Informational (consider updating)
- [ ] Issue description — file path — suggested fix

## Verified Current
- List of items confirmed as up-to-date

## Recommendations
- Prioritized list of updates to make

Save the report to progress/audit-report-[date].md if the user wants to track it.

Suggest 2-3 relevant next steps:

  • /lesson — to update specific curriculum content based on findings
  • /explain — to deep-dive into any new techniques discovered during the audit
  • /roadmap — to see if the learning path needs adjustment based on ecosystem changes

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.