This file governs all AI agent sessions in this repository.
- Repo: agentic-engineering
- Purpose: Umbrella documentation for the agentic engineering ecosystem
- License: MIT
- Status: Active development
- Type: Docs-only repo — no source code, no tests, no CI/CD
- Content language: English only
| Repo | Role | Path |
|---|---|---|
| ai-governance-framework | What may agents do? | ~/Github repos/ai-governance-framework |
| ai-project-management | What to do? | ~/Github repos/ai-project-management |
| ai-engineering-standards | How to do it? | ~/Github repos/ai-engineering-standards |
| ai-project-templates | Scaffolder | ~/Github repos/ai-project-templates |
- File names:
kebab-case.md - Directory names:
kebab-case/ - All content in English — no exceptions
- Cross-references use relative links:
[file](./path/file.md) - Diagrams use Mermaid syntax in fenced code blocks
- No placeholder text, no TODOs, no empty sections
- Read README.md — understand current content and structure
- Confirm scope with user: which doc, what outcome
- Do not create or modify files until scope is confirmed
- One document at a time
- After each file: verify cross-references resolve
- Keep README.md as the primary landing page — it must always be self-contained
- Summarize: files created/modified, cross-references updated
- Verify all internal links resolve to existing files
- Every file has real, usable content — no stubs
- README.md must work as a standalone introduction even without the docs/ folder
- Diagrams must render correctly in GitHub markdown
- Could a developer read this repo and understand the full ecosystem? If not, the content is incomplete.