
Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and Gemini CLI: local deterministic AST parsing, every edge explained, no vector store.
This skill helps an agent write developer docs from the actual working system instead of guesses. It separates implemented behavior from proposals, chooses the right doc form, and checks the finished Markdown with a bundled validator.
Builders who want Claude Code or Codex to create readable, accurate documentation from a live codebase.
You can produce docs that match current behavior and catch Markdown problems before you ship them.
`SKILL.md` defines how the host should read the system first, then write documentation that matches what it finds.
`references/documentation-types.md` and `references/writing-standard.md` explain which doc format to use and how to write it well.
`scripts/validate_docs.py` checks UTF-8, heading structure, fenced code blocks, local links, and image alt text.
`tests/test_validate_docs.py` exercises the validator without third-party Python packages.
`docs/quickstart.md`, `docs/reference.md`, and `docs/troubleshooting.md` cover installation, use, and common failures.
git clone https://github.com/NateBJones-Projects/developer-documentation-skill.git developer-documentation
$developer-documentation Create a quickstart and CLI reference for this project. Verify every documented command.
python3 developer-documentation/scripts/validate_docs.py PATH [PATH ...]
python3 -m unittest discover \ -s developer-documentation/tests \ -p "test_*.py"
developer-documentation is a standalone Agent Skills package for creating, restructuring, and maintaining accurate documentation for software, tools, automations, packages, APIs, SDKs, CLIs, installations, migrations, operations, and troubleshooting.
The skill reads the working system before it writes. It separates implemented behavior from proposals, chooses the documentation form that matches the reader's job, and checks the finished Markdown with a bundled validator.
Download developer-documentation-skill-1.0.1.zip from the latest release. The ZIP contains one complete developer-documentation directory that you can place in your AI host's skill directory.
To follow the project or contribute changes, clone the repository:
git clone https://github.com/NateBJones-Projects/developer-documentation-skill.git developer-documentation
Use the skill for:
developer-documentation/
├── ATTRIBUTION.md
├── LICENSE
├── README.md
├── SKILL.md
├── docs/
│ ├── quickstart.md
│ ├── reference.md
│ └── troubleshooting.md
├── references/
│ ├── documentation-types.md
│ └── writing-standard.md
├── scripts/
│ └── validate_docs.py
└── tests/
└── test_validate_docs.py
SKILL.md contains the working method loaded by the AI host. The docs directory explains how to install, invoke, and operate the skill. The references directory contains detailed writing guidance. The validator and its tests are independent Python files with no third-party dependencies.
Keep the complete developer-documentation directory together. Place it in the skill directory used by your AI host.
Common locations are:
| Host and scope | Destination |
|---|---|
| Codex project | .agents/skills/developer-documentation |
| Codex user | ~/.agents/skills/developer-documentation |
| Claude Code project | .claude/skills/developer-documentation |
| Claude Code user | ~/.claude/skills/developer-documentation |
Do not overwrite an existing directory with the same name until you confirm that it is obsolete. Restart the AI host if the newly installed skill does not appear.
Invoke the skill by name and state the documentation result you need. For example:
$developer-documentation Create a quickstart and CLI reference for this project. Verify every documented command.
The host can also select the skill automatically when the request clearly asks for developer documentation.
For a complete first run, see the quickstart.
Use the bundled validator when the project does not already have a documentation checker:
python3 developer-documentation/scripts/validate_docs.py PATH [PATH ...]
The validator checks UTF-8 decoding, heading structure, fenced code blocks, local links, and image alternative text. Add --check-fragments for GitHub-style heading targets. Add --allow-empty-alt only when every image with empty alternative text is confirmed to be decorative.
The validator does not prove that technical claims are correct or that commands work. The skill checks those against the current code, interface, tests, or runtime.
The documentation is licensed under CC BY 4.0. The validator, tests, and GitHub workflow are licensed under the MIT License. Both licenses permit commercial use.
Portions of the documentation adapt the Google developer documentation style guide. See ATTRIBUTION.md for the source, modification notice, and terms. This project is independent and is not endorsed by Google.
See LICENSE for the file-by-file license boundary.
From the directory that contains developer-documentation, run:
python3 -m unittest discover \
-s developer-documentation/tests \
-p "test_*.py"
For the full interface and validation behavior, see the skill reference. For common failures, see troubleshooting.
Sign in to join the discussion.
No comments yet. Be the first to say what this is good for.

Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and Gemini CLI: local deterministic AST parsing, every edge explained, no vector store.
Topic in, narrated explainer video out. A Claude Code / Codex skill that turns any topic into a black-canvas motion-graphics explainer video with TTS voiceover, subtitles and a chapter progress bar. Chinese or English; every frame drawn in code with Remotion.
Public repository for Agent Skills
Open-source AI job search: scan job portals, evaluate listings into a structured A-H report with a global 1-5 score, tailor your CV, track applications — runs locally in your AI coding CLI (Claude Code, Codex, OpenCode, Antigravity…)
Agent skill for beautiful, verifiable architecture, workflow, sequence, data-flow, and lifecycle diagrams—self-contained HTML with motion and crisp export.
A skill to stop your coding agent from burying the answer. ADHD-friendly output.