When should I run the llms-md-updater?

It is best used after any core logic change, provider addition, or before a release to ensure that AI assistants using your documentation have the most current information.

Does it support validating code examples?

Absolutely. One of its core functions is an Example Validator sub-agent that checks code snippets in your documentation against the actual implementation to prevent broken examples.

Can it update the capability matrix for different AI providers?

Yes, it specifically scans provider implementations to update matrices that define which features (like streaming or tool calling) are supported by which model.

How does this skill detect changes in my codebase?

It uses Git logs to identify modified files since the last documentation update and deploys specialized analysis agents to scan source code, types, and exported APIs.

What is the purpose of an llms.md file?

An llms.md file is a specialized documentation standard optimized for Large Language Models, providing them with the necessary context, API patterns, and constraints to generate accurate code for a specific project.

LLMs.md Documentation Updater

Name: LLMs.md Documentation Updater
Author: ProviderProtocol

byProviderProtocol

•

학습 및 문서화

Synchronizes and optimizes AI-ready documentation by automatically updating llms.md files based on codebase changes.

The llms-md-updater is a specialized capability designed to eliminate documentation drift in AI-first development environments. It automates the maintenance of llms.md files—crucial reference guides that provide LLMs with the context needed to understand your codebase. By leveraging sub-agents to scan for new providers, interface changes, and API updates, this skill ensures that your technical documentation always reflects the ground truth of your implementation. It is particularly effective for SDK maintainers who need to keep capability matrices and code examples accurate for both human developers and AI assistants.

주요 기능

01LLM-optimized formatting for high-fidelity context ingestion

02Intelligent capability matrix generation for multi-provider architectures

03Source-to-doc synchronization for providers, types, and public API surfaces

04Automated validation and correction of code examples against current implementation

05Automated codebase analysis using Git history and parallel sub-agents

063 GitHub stars

사용 사례

01Syncing the LLM reference guide after major type refactors or API signature changes

02Updating documentation immediately after adding new AI providers or model capabilities

03Ensuring production-ready documentation accuracy before repository releases

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add providerprotocol/ai llms-md-updater

For use in Claude.ai and ChatGPT

Download Skill