Inspect a repository and draft an AGENTS.md file using the standard template, capturing commands, structure, and workflow rules.
Installation
Details
Usage
After installing, this skill will be available to your AI coding assistant.
Verify installation:
skills listSkill Instructions
name: agents-md-assistant description: Inspect a repository and draft an AGENTS.md file using the standard template, capturing commands, structure, and workflow rules.
AGENTS.md Assistant
Quick start
- Review
README.md,CONTRIBUTING.md, and docs/CI configs to learn the intended workflow. - Pull exact commands from build scripts (
Makefile,package.json,pyproject.toml,justfile). - Map key directories (source, tests, docs, scripts) and mark generated or vendor paths.
- Capture architecture: key entrypoints, components, and configuration locations.
- Fill unknowns with TODOs and confirm them with the user.
Workflow
-
Inspect the repository
- Read
README.md,CONTRIBUTING.md, and/docsfor goals and workflows. - Check build/config files:
Makefile,package.json,pyproject.toml,justfile,Dockerfile, CI pipelines.
- Read
-
Extract commands
- Record install, dev, lint, format, typecheck, test, and build commands exactly as documented.
- If multiple options exist, capture the preferred default and note alternates.
-
Map structure
- Identify primary code, tests, docs, scripts, and generated/ignored folders.
- Note where key configs live and any build artifacts to avoid editing.
-
Summarize architecture & environment
- Document key components, data flow, and entrypoints.
- Capture required versions, services, env vars, and migration/seed steps.
-
Author AGENTS.md
- Use
references/agents_md_template.mdas the base. - Place
AGENTS.mdat the repo root unless instructed otherwise. - Flag missing info with TODOs and ask for confirmation.
- Use
Reference
references/agents_md_template.md: base template to fill.
More by dmitriiweb
View allWrite Python docstrings following the Google Python Style Guide, using clear sections and examples.
Write focused pytest tests as standalone functions (one test per function), avoiding test classes.
Run the project's formatter, linters, and mypy checks in the required order, fixing issues and managing any needed stub dependencies via uv.
Perform structured, actionable code reviews for Python code with clear findings and suggestions.
