Skip to content
Back to skills

Llm Doc

ASecurity

Generate the LLM-friendly API cheat sheets in docs_llm/api from Doxygen XML. Use when public headers or their Doxygen comments changed and docs_llm/api must be refreshed.

  • 14 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added November 7, 2026
ai-agentspythonapi

Works with

  • api

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned November 7, 2026

npx -y skills add ezMbedded/EasyEmbeddedFramework --skill llm-doc --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Llm Doc?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Llm Doc
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/ezmbedded-llm-doc/badge)](https://www.skillsdirectory.com/skills/ezmbedded-llm-doc)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: llm-doc
description: Generate the LLM-friendly API cheat sheets in docs_llm/api from Doxygen XML. Use when public headers or their Doxygen comments changed and docs_llm/api must be refreshed.
---

# Generate LLM-friendly API cheat sheets

## Purpose
Generate one LLM-friendly API cheat sheet per module, which section 5 of `docs_llm/LLM_GUIDE.md` points to.

## When to use
- A public header under `inc/` or its Doxygen comments changed.

## Inputs
- Doxygen XML directory: `doxygen/xml` (script default)
- Output directory: `docs_llm/api` (script default)
- Public header root for the coverage check: `inc` (script default)

## Prerequisites
- Doxygen and graphviz installed; the Doxyfile has `GENERATE_XML = YES`, `XML_PROGRAMLISTING = YES` and the `isr_safe` / `blocking` aliases.
- Headers are documented. Run the code-document skill first; undocumented headers produce thin cheat sheets.

## Steps
1. Build the XML from the repository root: `.claude/skills/code-document/scripts/run_doxygen.sh`. If it exits with 2 (Doxygen missing or failed), stop and report it. Warnings (exit 1) do not block this skill, but list them in the report.
2. Run `python3 .claude/skills/llm-doc/scripts/gen_llm_doc.py <header>`, for example `inc/service/event_bus/ez_event_bus.h`. Only the given header(s) are parsed and written, and `INDEX.md` is updated for it while the other modules stay as they are. Without a header argument every public header is processed; do that only when the user asks for all modules. The defaults above apply; pass `--xml`, `--out` or `--inc` only to override them.
3. Read the script output. Exit code 1 means a header failed to process (`FAILED`) or, on a full run, headers under `inc/` have no cheat sheet (`MISSING`); the script lists them. Exit code 2 is a fatal error, for example the header is not in the Doxygen XML.
4. Open a few generated `<module>.md` files and check them: signature present, `Parameters` and `Return values` not empty, `ISR-safe:` and `Blocking:` lines present for functions. An empty section means the Doxygen comment in the header is missing or wrong.
5. Confirm section 5 of `docs_llm/LLM_GUIDE.md` points to `docs_llm/api/`. Do not edit the generated files to fix content.

## Output files
| File | Content |
|------|---------|
| `docs_llm/api/<module>.md` | Cheat sheet per public header (`.h` and `.hpp`) |
| `docs_llm/api/INDEX.md` | Module table: module, header, purpose, function count |

## Rules
- Generated files are never edited by hand; fix the Doxygen comment in the header and regenerate.
- Do not change the script's output format without checking `LLM_GUIDE.md` and `AGENTS.md`, which refer to the literal text `ISR-safe: yes`.

## Report back
- Modules generated.
- Modules failed or missing (from the script's `FAILED` and `MISSING` lists).
- Modules not generated correctly (from step 4), with the header and symbol.
- Any errors or warnings from Doxygen or the script.

Files in this skill

  • SKILL.md2.9 KB
  • scripts/gen_llm_doc.py17.7 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.