agent-discipline
Skill1 starsChanged 41 days agoMITView on GitHub
Prevent over-engineering, silent assumptions, and unrelated edits. Use when coding any task, especially when agents sprawl or invent requirements.
In this repository49 other files it gives its agents.
- accessibilitySkill
- address-feedbackSkill
- agent-handoffSkill
- api-designSkill
- architecture-decisionSkill
- brainstormSkill
- browser-verifySkill
- code-reviewSkill
- containerizeSkill
- context-efficientSkill
- create-prSkill
- database-changesSkill
- debug-systematicSkill
- dependency-upgradeSkill
- docs-groundedSkill
- error-handlingSkill
- explore-codebaseSkill
- fix-ciSkill
- frontend-designSkill
- git-commitSkill
- git-worktreeSkill
- incident-triageSkill
- incremental-slicesSkill
- legacy-changeSkill
- migrate-codeSkill
- minimal-diffSkill
- observabilitySkill
- parallel-workSkill
- performance-auditSkill
- plan-before-codeSkill
- python-qualitySkill
- react-componentsSkill
- refactor-safeSkill
- release-notesSkill
- security-reviewSkill
- setup-ciSkill
- shell-safetySkill
- ship-checklistSkill
- simplify-codeSkill
- skill-authoringSkill
- skill-routerSkill
- spec-drivenSkill
- tddSkill
- threat-modelSkill
- triage-issuesSkill
- typescript-strictSkill
- verify-workSkill
- write-docsSkill
- write-testsSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 10
Do not invent: if unsure, read code or ask.
- Distinctive: line 12
No speculative architecture: no extra layers, configs, or “future-proofing” for hypothetical needs.
- Distinctive: line 15
Verify: run checks; do not claim done on vibes.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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.
Files
Agent Discipline
Rules
- Do not invent: if unsure, read code or ask; do not guess product behavior.
- Minimal change: touch only what the task requires; no unrelated cleanup.
- No speculative architecture: no extra layers, configs, or “future-proofing” for hypothetical needs.
- Match the repo: reuse patterns, libraries, and naming already present.
- Surfaces assumptions: state them explicitly when they affect the design.
- Verify: run checks; do not claim done on vibes.
Anti-Patterns
- Rewriting a working file for style
- Adding dependencies for trivial helpers
- Expanding scope mid-task without user OK
- Silent API/behavior changes
Self-Check
- Diff matches the request
- No drive-by files
- Assumptions listed if any
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.