coding-kit
CLAUDE.md3 starsChanged 11 days agoMITView on GitHub
AGENTS.md points to OPS.md, the full contract. This file only adapts it to a run on a GitHub-hosted runner, where the machine-local parts cannot apply. You run from .github/workflows/claude.yml on ubuntu-latest, triggered by an @claude mention in an issue or pull request comment. - The memory store (~/.memory, its search and findings tools, the warmup script) does not exist on the runner. Skip memory search and saving (OPS §5) and say so in one line instead of failing the step.…
In this repository55 other files it gives its agents.
- AGENTS.mdAGENTS.md
- architecture-simplicitySkill
- autonomous-workSkill
- brainstormingSkill
- code-graph-reviewSkill
- code-review-and-qualitySkill
- dashboard-designSkill
- design-systemSkill
- dev-wikiSkill
- dispatching-parallel-agentsSkill
- fable-judgeSkill
- fable-methodSkill
- git-workflow-and-versioningSkill
- money-path-safetySkill
- observability-and-instrumentationSkill
- ponytailSkill
- production-first-decisionsSkill
- receiving-code-reviewSkill
- security-and-hardeningSkill
- shipping-and-launchSkill
- skill-authoringSkill
- systematic-debuggingSkill
- testing-disciplineSkill
- using-git-worktreesSkill
- verification-before-completionSkill
- web-researchSkill
- windows-encoding-fixesSkill
- writing-plansSkill
- architecture-simplicitySkill
- autonomous-workSkill
- brainstormingSkill
- code-graph-reviewSkill
- code-review-and-qualitySkill
- dashboard-designSkill
- design-systemSkill
- dev-wikiSkill
- dispatching-parallel-agentsSkill
- fable-judgeSkill
- fable-methodSkill
- git-workflow-and-versioningSkill
- money-path-safetySkill
- observability-and-instrumentationSkill
- ponytailSkill
- production-first-decisionsSkill
- receiving-code-reviewSkill
- security-and-hardeningSkill
- shipping-and-launchSkill
- skill-authoringSkill
- systematic-debuggingSkill
- testing-disciplineSkill
- using-git-worktreesSkill
- verification-before-completionSkill
- web-researchSkill
- windows-encoding-fixesSkill
- writing-plansSkill
The file
We split this CLAUDE.md into the rules it gives an agent
2 rules. About 0 are common, 0 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 22
Run checks matching the affected contract.
- Very specific: line 41
Encoding: Every subprocess.* call with text=True also passes encoding="utf-8".
≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.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
Claude Code
@AGENTS.md
AGENTS.md points to OPS.md, the full contract. This file only adapts it to a run on a GitHub-hosted runner, where the machine-local parts cannot apply.
Runner context
You run from .github/workflows/claude.yml on ubuntu-latest, triggered by an
@claude mention in an issue or pull request comment.
- The memory store (
~/.memory, its search and findings tools, the warmup script) does not exist on the runner. Skip memory search and saving (OPS §5) and say so in one line instead of failing the step. - There is no
.override.mdhere, so the default contract applies. - Report model or tool limitations that affect verification; do not attribute unmeasured behavior to a provider.
Verifying a change
Run checks matching the affected contract. python -m pytest tests -q is the
repository-wide check when shared policy or runtime changes warrant it;
focused checks suffice for isolated changes. scripts/doctor.py diagnoses
installation health, not model behavior.
For an installation task on a disposable runner, install prerequisites, run
scripts/install.py, then the doctor. Do not bootstrap or mutate a real user
memory root to verify an unrelated edit. Report environmental failures
separately from failures caused by the change.
Repository gates to respect before proposing a push:
- File size. Code soft 500 / hard 1000 lines, docs soft 300 / hard 500.
Split rather than grow; never edit
scripts/tools/file_size_baseline.jsonto make the check pass. - Integrity manifest.
integrity-manifest.jsonhash-pins the control plane. After changing a file in its scope, regenerate it withpython scripts/tools/integrity_manifest.py --updatefor your own change only, never to hide someone else's drift. - Encoding. Every
subprocess.*call withtext=Truealso passesencoding="utf-8"; an AST-based gate fails the build otherwise.
Conventions
- English for code, comments and commit messages.
- Conventional commits matching the log:
fix(policy): ...,docs: ...,test: .... - Commits, pushes and pull requests follow OPS §1. When publication is
authorized, use a branch and PR rather than pushing to
master. - Report exact commands, checked state and coverage; never claim broader coverage than was exercised.
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.