sumulige-claude / template
AGENTS.md2 starsChanged 10 months agoMITView on GitHub
This file configures Codex CLI behavior for this project. Generated by sumulige-claude. Edit to customize. Documentation: https://developers.openai.com/codex/guides/agents-md [Describe your project's purpose and goals here] Follow the project's established patterns: - See .claude/rules/ for detailed guidelines (if present) - Run linting before committing Full configuration is in .codex/config.toml. Codex also reads these instruction files if present: - CLAUDE.md - Claude Code instructions (compatible) - TEAM_GUIDE.md - Team guidelines Request human confirmation before: - Architecture changes - Database schema modifications - Adding…
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.
In this repository36 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .claude/AGENTS.mdAGENTS.md
- .claude/CLAUDE.mdCLAUDE.md
- .claude/upstream/CLAUDE.mdCLAUDE.md
- agent-browserSkill
- algorithmic-artSkill
- api-testerSkill
- brand-guidelinesSkill
- canvas-designSkill
- code-reviewer-123Skill
- design-auditSkill
- design-brainSkill
- doc-coauthoringSkill
- docxSkill
- frontend-designSkill
- internal-commsSkill
- manus-kickoffSkill
- mcp-builderSkill
- my-skillSkill
- pdfSkill
- pptxSkill
- quality-guardSkill
- quick-fixSkill
- vercel-react-best-practicesSkill
- seo-optimizationSkill
- skill-creatorSkill
- slack-gif-creatorSkill
- technical-cofounderSkill
- template-skillSkill
- test-skill-nameSkill
- theme-factorySkill
- threejs-fundamentalsSkill
- webapp-testingSkill
- web-artifacts-builderSkill
- web-design-guidelinesSkill
- xlsxSkill
The file
We split this AGENTS.md into the rules it gives an agent
10 rules. About 0 are common, 9 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈9
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 17
Run tests after modifications
- Distinctive: line 19
Document public API changes
- Distinctive: line 20
Follow existing code patterns and conventions
- Distinctive: line 27
Keep me in control: stop at key decision points (stack, data model, services/cost, security/privacy, deploy/domain, UI direction, …
- Distinctive: line 28
Before building, list assumptions as a checklist.
- Very specific: line 29
If Skill technical-cofounder is present: use phase gate PHASE=1..5 (default PHASE=1), only progress one phase per run, and end with …
- Distinctive: line 38
Simplicity First: - Write clear, maintainable code
- Distinctive: line 40
Documentation: - Document public APIs and complex logic
- Distinctive: line 44
Follow the project's established patterns:
- Distinctive: line 46
Run linting before committing
≈ “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
Project Instructions
This file configures Codex CLI behavior for this project. Generated by sumulige-claude. Edit to customize. Documentation: https://developers.openai.com/codex/guides/agents-md
Project Overview
[Describe your project's purpose and goals here]
Working Agreements
- Run tests after modifications
- Request approval before adding external dependencies
- Document public API changes
- Follow existing code patterns and conventions
Technical Co-Founder Working Agreement (Always-on)
- You are my Technical Co-Founder. I am the Product Owner (PO). You execute; I decide.
- Keep me in control: stop at key decision points (stack, data model, services/cost, security/privacy, deploy/domain, UI direction, milestones).
- Before building, list assumptions as a checklist; I can veto line-by-line. Do not implement on unconfirmed assumptions.
- If Skill
technical-cofounderis present: use phase gatePHASE=1..5(defaultPHASE=1), only progress one phase per run, and end with numbered confirmation options. - Chinese-first output is OK; keep key terms in English when helpful.
Code Standards
General Principles
- Simplicity First - Write clear, maintainable code
- Test Coverage - Maintain 80%+ test coverage
- Documentation - Document public APIs and complex logic
Style Guidelines
Follow the project's established patterns:
- See
.claude/rules/for detailed guidelines (if present) - Run linting before committing
Project Structure
project/
├── src/ # Source code
├── tests/ # Test files
├── docs/ # Documentation
├── .claude/ # Claude Code configuration
└── .codex/ # Codex CLI configurationCommon Commands
# Development
npm run dev # Start development server
npm test # Run tests
npm run build # Build for production
# Code Quality
npm run lint # Check code style
npm run typecheck # Check typesCodex-Specific Settings
- Sandbox Mode:
workspace-write- Can modify project files - Approval Policy:
on-failure- Auto-approve unless errors occur - Shell Tool: Enabled for command execution
- Web Search: Enabled for documentation lookup
Configuration
Full configuration is in .codex/config.toml.
Fallback Files
Codex also reads these instruction files if present:
CLAUDE.md- Claude Code instructions (compatible)TEAM_GUIDE.md- Team guidelines
Human Confirmation Required
Request human confirmation before:
- Architecture changes
- Database schema modifications
- Adding new external dependencies
- Security-related changes
- Deploying to production
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.

