skills
AGENTS.md1 starsChanged 7 months agoApache-2.0View on GitHub
This project is indexed by GitNexus as skills (204 symbols, 264 relationships, 5 execution flows). Use the GitNexus MCP tools to understand code, assess impact, and navigate safely. If any GitNexus tool warns the index is stale, run npx gitnexus analyze in terminal first.
Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 31,439 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
The file
Show its rules and text
We split this AGENTS.md into the rules it gives an agent
15 rules. About 0 are common, 2 are distinctive, and 13 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
13
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 7
Treat local runtime state as noise, not source: .omx/, .sisyphus/, .gitnexus/, and .claude/settings.local.json are ignored/local artifacts.
- Very specific: line 21
Do not write wiki/ pages until the target wiki language is explicit in the request or already remembered in wiki/config.md.
- Very specific: line 36
If GitNexus reports a stale index, refresh it with npx gitnexus analyze before trusting graph results.
- Very specific: line 37
Before editing a function, class, or method, run gitnexus_impact({target: "symbolName", direction: "upstream"}) and surface HIGH/CRITICAL …
- Very specific: line 38
Before committing, run gitnexus_detect_changes() to confirm the change scope matches expectations.
- Very specific: line 39
Use gitnexus_rename for symbol renames.
- Very specific: line 45
Verify the smallest relevant surface instead: for llm-wiki, use validate_skill.py.
- Very specific: line 56
MUST run impact analysis before editing any symbol: Before modifying a function, class, or method, run gitnexus_impact({target: …
- Distinctive: line 58
MUST warn the user: if impact analysis returns HIGH or CRITICAL risk before proceeding with edits.
- Very specific: line 59
When exploring unfamiliar code, use gitnexus_query({query: "concept"}) to find execution flows instead of grepping.
- Very specific: line 60
When you need full context on a specific symbol — callers, callees, which execution flows it participates in — use gitnexus_context({name: …
- Very specific: line 64
NEVER edit a function, class, or method without first running gitnexus_impact on it.
- Distinctive: line 65
NEVER ignore HIGH or CRITICAL risk warnings from impact analysis.
- Very specific: line 66
NEVER rename symbols with find-and-replace — use gitnexus_rename which understands the call graph.
- Very specific: line 67
NEVER commit changes without running gitnexus_detect_changes() to check affected scope.
≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md 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.
Files
Agent Skills Repo Notes
Repo shape
- This repo is a skill library, not an app or package workspace. There is no
package.json, lockfile, CI workflow, Makefile, or general build/test/lint pipeline at the repo root. - Each skill should live in a top-level folder with
SKILL.mdas the entry point.references/,samples/, andscripts/are optional support folders. - Treat local runtime state as noise, not source:
.omx/,.sisyphus/,.gitnexus/, and.claude/settings.local.jsonare ignored/local artifacts.
Highest-value files
README.md/docs/README.zh-CN.md— repo purpose and the intended top-level skill layout.llm-wiki/SKILL.md— canonical workflow for the main published skill in this repo.llm-wiki/references/wiki-schema.md— the durable page schema and language/provenance rules behindllm-wiki.llm-wiki/scripts/install_skill.py— the real install behavior forcodex,claude-code, andgeneric-agenttargets.llm-wiki/scripts/validate_skill.py— the closest thing this repo has to a focused verification command..claude/skills/gitnexus/*/SKILL.md— local guidance for GitNexus-backed exploration, impact analysis, debugging, refactoring, and CLI usage.
llm-wiki rules that are easy to miss
raw/is immutable source evidence. Never edit, move, delete, or generate files under it.- Do not write
wiki/pages until the target wiki language is explicit in the request or already remembered inwiki/config.md. - For large ingests,
wiki/manifest.jsonis required.llm-wikitreats more than 30 raw files as batch mode, with 20 files per batch by default. - Source-page generation and concept synthesis are intentionally separated in
llm-wiki/SKILL.md; do not collapse them into one undifferentiated pass. wiki/index.mdis a compact navigation page, not the full provenance inventory. Full raw coverage belongs inwiki/manifest.json.
Commands agents are likely to guess wrong
- Install the published skill via npm wrapper:
npx skills@latest add inkpark/skills/llm-wiki - Preview a local install without writing:
python3 llm-wiki/scripts/install_skill.py --dry-run --platform codex - Real local install requires an explicit confirmation flag:
python3 llm-wiki/scripts/install_skill.py --platform claude-code --confirm generic-agentinstalls need--targetorLLM_WIKI_SKILL_TARGET/AI_AGENT_SKILL_TARGET; there is no default generic skill path.- For
llm-wikichanges, run the focused validator:python3 llm-wiki/scripts/validate_skill.py --repo-root /home/inkparker/workspace/skills
GitNexus usage in this repo
- If GitNexus reports a stale index, refresh it with
npx gitnexus analyzebefore trusting graph results. - Before editing a function, class, or method, run
gitnexus_impact({target: "symbolName", direction: "upstream"})and surface HIGH/CRITICAL risk before proceeding. - Before committing, run
gitnexus_detect_changes()to confirm the change scope matches expectations. - Use
gitnexus_renamefor symbol renames; do not do graph-blind find/replace. - For unfamiliar code, prefer
gitnexus_query/gitnexus_contextover blind grep when the index is fresh.
Practical verification expectations
- There is no repo-wide build/test command to run after every change.
- Verify the smallest relevant surface instead: for
llm-wiki, usevalidate_skill.py; for instruction-only edits, verify by rereading the touched files against the cited source files.
<!-- gitnexus:start -->GitNexus — Code Intelligence
This project is indexed by GitNexus as skills (204 symbols, 264 relationships, 5 execution flows). Use the GitNexus MCP tools to understand code, assess impact, and navigate safely.
If any GitNexus tool warns the index is stale, run
npx gitnexus analyzein terminal first.
Always Do
- MUST run impact analysis before editing any symbol. Before modifying a function, class, or method, run
gitnexus_impact({target: "symbolName", direction: "upstream"})and report the blast radius (direct callers, affected processes, risk level) to the user. - MUST run
gitnexus_detect_changes()before committing to verify your changes only affect expected symbols and execution flows. - MUST warn the user if impact analysis returns HIGH or CRITICAL risk before proceeding with edits.
- When exploring unfamiliar code, use
gitnexus_query({query: "concept"})to find execution flows instead of grepping. It returns process-grouped results ranked by relevance. - When you need full context on a specific symbol — callers, callees, which execution flows it participates in — use
gitnexus_context({name: "symbolName"}).
Never Do
- NEVER edit a function, class, or method without first running
gitnexus_impacton it. - NEVER ignore HIGH or CRITICAL risk warnings from impact analysis.
- NEVER rename symbols with find-and-replace — use
gitnexus_renamewhich understands the call graph. - NEVER commit changes without running
gitnexus_detect_changes()to check affected scope.
Resources
| Resource | Use for |
gitnexus://repo/skills/context | Codebase overview, check index freshness |
gitnexus://repo/skills/clusters | All functional areas |
gitnexus://repo/skills/processes | All execution flows |
gitnexus://repo/skills/process/{name} | Step-by-step execution trace |
CLI
| Task | Read this skill file |
| Understand architecture / "How does X work?" | .claude/skills/gitnexus/gitnexus-exploring/SKILL.md |
| Blast radius / "What breaks if I change X?" | .claude/skills/gitnexus/gitnexus-impact-analysis/SKILL.md |
| Trace bugs / "Why is X failing?" | .claude/skills/gitnexus/gitnexus-debugging/SKILL.md |
| Rename / extract / split / refactor | .claude/skills/gitnexus/gitnexus-refactoring/SKILL.md |
| Tools, resources, schema reference | .claude/skills/gitnexus/gitnexus-guide/SKILL.md |
| Index, status, clean, wiki CLI commands | .claude/skills/gitnexus/gitnexus-cli/SKILL.md |
<!-- gitnexus:end -->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
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.

