Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    API Documentation Generator — Docs That Match The Code

    1

    Generate code-accurate API references and gap reports by joining the exported surface with existing documentation.

    $12

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    API Documentation Generator — Docs That Match The Code

    API Documentation Generator — Docs That Match The Code

    Example session with this skill installed

    Generate an API reference for the packages/core directory and list any undocumented exports.

    • Read your context and instructions
    • Compiled the api documentation generator

    I have analyzed packages/core. I identified 12 exported symbols across 4 files and compared them against README.md and docs/api.md.

    API Documentation Analysis: packages/core

    The Package

    • Path: packages/core
    • Surface: 12 symbols (8 in barrel, 4 deep-import only)
    • Corpus: README.md, docs/api.md

    The Reference

    StorageClient (class)

    • Source: src/client.ts
    • Signature: export class StorageClient extends BaseClient
    • Doc Comment: "Primary entry point for the storage provider API."

    upload(config: Config): Promise<Result> (function)

    • Source: src/upload.ts
    • Signature: export async function upload(config: Config): Promise<Result>
    • Description: Quoted from body: Handles multi-part uploads with retry logic.

    The Gap List
    The following symbols are exported but not mentioned in any documentation:

    • validateToken: (function) src/auth.ts
    • RetryStrategy: (interface) src/types.ts
    • BACKOFF_FACTOR: (constant) src/constants.ts

    The Drift List

    • LegacyUploader: Mentioned in README.md (line 42) but no longer exported. Replaced by StorageClient.

    What the reference does not cover

    • Internal utility functions in src/utils/ (non-exported).
    • Environmental requirements for the Config object.

    Final Checklist

    • All signatures quoted from source.
    • Deep exports marked.
    • Drift identified.

    Next steps

    1. Review the Gap List to decide if validateToken should be documented or made private.
    2. Update README.md line 42 to point to StorageClient instead of the defunct LegacyUploader.
    3. Move RetryStrategy into the barrel if it is intended for public consumption.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Documentation generators often hallucinate parameters or miss renamed exports, resulting in beautiful prose that is factually wrong. This skill solves the documentation-drift problem by performing a read-only analysis of your actual source code and barrel files. It identifies exactly what is exported, what is missing from your documentation corpus, and what your current docs mention that no longer exists in the code.

    What it does

    • Exported surface analysis identifies every exported symbol, its kind, and its source file to map the package's true interface.
    • Gap detection reports public symbols that are present in the code but completely missing from your documentation.
    • Drift identification flags symbols mentioned in your README or docs that have been renamed or removed from the source.
    • Signature extraction quotes exact code signatures and doc comments directly from the source to prevent hallucinated behaviors.
    • Barrel verification compares the internal exported surface against the public barrel to find deep-reachable symbols.

    How it works

    1. Frame the package by identifying the target directory, the intended audience, and any existing documentation files.
    2. Analyze the surface using agent-context or git to extract symbols, kind, and external dependencies.
    3. Map the corpus by searching your repository for Markdown and text files that serve as the current documentation.
    4. Join and report by checking every export against the corpus to generate the gap list and the final API reference document.

    Frameworks & tools

    This tool works with any repository using git and standard text-based documentation formats like .md, .mdx, or .rst. It integrates with the ai-agent-context CLI for deeper JSON-based analysis of module responsibilities and team decisions.

    Why this beats prompting it yourself

    Standard prompts rely on the LLM's memory of your files, leading to missed exports and outdated signatures. This skill enforces a "code-first" join that explicitly lists documented symbols that don't exist, ensuring 100% alignment between your code and your reference.

    Use cases

    • Generate a complete API reference for a new or undocumented library.
    • Audit existing READMEs to find functions or classes missing from the docs.
    • Repair documentation after a major refactor or symbol renaming.
    • Onboard new maintainers by providing a structural map of the public API.

    Known limitations

    This is a read-only analysis skill that generates documents in the chat; it does not write or overwrite files in your repository. It does not run tests, execute code, or perform builds.

    Read-only analysis only. Does not write to files. Requires manual inspection of barrel files. CLI analysis is optional but degrades to standard git/grep without ai-agent-context.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 27 days ago

    Requires

    ai-agent-context
    Node.js
    agent-context
    Git

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Network Access
    Listed27 days ago

    What's inside

    Frequently Asked Questions

    Unlike standard LLMs that rely on memory, this skill performs a code-first join by cross-referencing your actual source exports against your Markdown files to flag missing, renamed, or non-existent symbols.

    The skill is language-agnostic for any repository using Git and text-based documentation (.md, .mdx, .rst), though it is optimized for environments where exported symbols can be clearly parsed from source or barrel files.

    No, this is a read-only analysis tool; it generates the documentation and gap reports within the chat interface for you to review and manually commit to your repository.

    You receive the full skill package including the logic for surface analysis, gap detection, and signature extraction, compatible with AI agents that support repository-level context tools.

    You need a Git-enabled repository and an AI agent capable of file-system access; for enhanced structural analysis, it integrates seamlessly with the ai-agent-context CLI.

    The 'Barrel Verification' feature audits your public-facing entry points to ensure that deep-reachable symbols are correctly exposed and documented, preventing broken imports for your users.

    New in Technical Documentation