FixLab
AGENTS.md0 starsChanged 59 days agoMITView on GitHub
FixLab is a reusable Agency Copilot workflow for turning defects and small enhancements into evidence-backed pull requests. Keep the core product-neutral: repository-specific paths, commands, ports, environments, and browser journeys belong in repository profiles and examples. 1. Read README.md, docs/architecture.md, the nearest AGENTS.md, and any matching .github/instructions/*.instructions.md learned rules. 2. Inspect git status and preserve unrelated worktree changes. 3. Reproduce behavioral defects or define a measurable acceptance check for small enhancements. 4. Make the smallest complete change. 5. Add focused regression…
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 31,774 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.
In this repository5 other files it gives its agents.
- dashboard/AGENTS.mdAGENTS.md
- .github/copilot-instructions.mdCopilot instructions
- codeblend-ai-compositeSkill
- fixlab-job-scoped-post-approvalSkill
- fixlab-windows-background-probesSkill
The file
We split this AGENTS.md into the rules it gives an agent
8 rules. About 1 is common, 4 are distinctive, and 3 are very specific to this project.
≈1
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- could be automatic: a setting could enforce it
- Very specific: line 12
Read README.md, docs/architecture.md, the nearest AGENTS.md, and any matching .github/instructions/*.instructions.md learned rules.
- Distinctive: line 18
Add focused regression coverage in the existing test suite.
- Common: line 24
Never commit credentials, tokens, cookies, browser state, customer data, or private endpoints.
- Very specific: line 28
Stop only processes started and tracked by the current FixLab job.
- Distinctive: line 29
Never select production automatically or bypass authentication, deployment, review, or data-safety controls.
- Distinctive: line 31
Do not use destructive Git operations without explicit approval.
- Distinctive: line 32
Do not report a skipped check as successful.
- Very specific: line 42
Use dependency-free Node.js APIs unless a dependency provides material, reviewable value.
≈ “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
FixLab agent instructions
Repository purpose
FixLab is a reusable Agency Copilot workflow for turning defects and small enhancements into evidence-backed pull requests. Keep the core product-neutral: repository-specific paths, commands, ports, environments, and browser journeys belong in repository profiles and examples.
Required workflow
- Read
README.md,docs/architecture.md, the nearestAGENTS.md, and any matching.github/instructions/*.instructions.mdlearned rules. - Inspect
git statusand preserve unrelated worktree changes. - Reproduce behavioral defects or define a measurable acceptance check for small enhancements.
- Make the smallest complete change.
- Add focused regression coverage in the existing test suite.
- Run
npm run validate. - Report every failed, skipped, timed-out, or blocked gate explicitly.
Safety invariants
- Never commit credentials, tokens, cookies, browser state, customer data, or private endpoints.
- Bind local services to loopback unless a reviewed design explicitly requires another interface.
- Stop only processes started and tracked by the current FixLab job.
- Never select production automatically or bypass authentication, deployment, review, or data-safety controls.
- Do not use destructive Git operations without explicit approval.
- Do not report a skipped check as successful.
Change boundaries
- CLI behavior belongs in
bin/. - Dashboard HTTP and job orchestration belong in
dashboard/. - Reusable agent behavior belongs in
agents/andprompts/. - Repository onboarding examples belong in
examples/andtemplates/. - Operational and security contracts must be reflected in
docs/.
Use dependency-free Node.js APIs unless a dependency provides material, reviewable value. Keep commands cross-platform where practical and use repository-owned scripts instead of undocumented shell snippets.
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.