Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Tool Discovery

    1

    Find and rank software tools based on specific constraints, live repository data, and functional requirements.

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Tool Discovery

    Tool Discovery

    Example session with this skill installed

    Find 4 open-source tools to convert OneNote files to Markdown. Must run on Linux CLI, work offline, and be MIT or Apache-2.0 licensed. include 'onenote-md-exporter' in the search.

    • Read your context and instructions
    • Compiled the tool discovery

    I have evaluated 5 candidates for OneNote to Markdown conversion based on your Linux CLI and offline requirements.

    Job

    Convert OneNote to Markdown. platform=linux must_cli=yes as_of=2024-05-20

    Recommendation: onenote-parser

    Why: This is the only candidate that parses the .one binary format directly in a Linux environment without requiring a Windows-only COM interface or a Microsoft Graph API connection. It is actively maintained with a recent commit in April 2024.
    Cite: https://github.com/example/onenote-parser

    Ranked table

    RankToolStarsLast commitPlatformCLILicenseNot for
    1onenote-parser8402024-04-12Cross-platformYesMITCloud-only notes
    2convert-one1.2k2023-11-05Linux/macOSYesApache-2.0Batch processing
    3onenote-md-exporter1502022-01-10Windows/LinuxYesMITProduction stability
    4note-migrate-gui2.1k2024-05-01Linux (Electron)NoGPL-3.0Headless/Cron jobs

    Also considered

    • ms-onenote-python: Eliminated. Requires active Microsoft Azure tenant and internet connection for Graph API.
    • one-to-md-win: Eliminated. Requires Windows COM+ and local OneNote installation.

    Sources

    Next steps

    1. Review the onenote-parser README for specific .one version compatibility.
    2. Verify if your local files are the newer package format or legacy binary files.
    3. Test the CLI output against a sample .one file to check for image embedding support.

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

    About this skill

    The problem

    Finding the right tool for a specific engineering task usually involves wading through stale blog posts, affiliate-heavy roundups, and misleading star counts. Developers often waste hours trial-installing software only to find it lacks a CLI or requires a GUI on a headless server.

    What it does

    • Performs multi-vector searches across code hosts and the open web to find candidates for a specific job.
    • Extracts structured metadata for every tool, including license type, last commit date, and platform support.
    • Enforces hard filters for constraints like offline capability, CLI availability, and specific license requirements.
    • Generates a ranked comparison table that prioritizes functional fit and maintenance health over raw popularity.
    • Identifies "not for" scenarios for every tool to prevent architectural mismatches.

    Why this beats prompting it yourself

    Generic LLM prompts often hallucinate features or rely on training data that misses recent repository activity. This skill forces the agent to fetch live repository metadata, check actual license files, and verify the last commit date before making a recommendation.

    Use cases

    • Finding an open-source, MIT-licensed CLI tool for a specific file conversion.
    • Comparing cross-platform system monitors that don't require a cloud account.
    • Researching migration paths between two technologies for a headless environment.
    • Shortlisting libraries based on active maintenance and specific language runtimes.

    Known limitations

    Requires an agent with HTTP access and web search capabilities. Star counts and commit dates are point-in-time snapshots based on the provided date stamp.

    Requires HTTP access and web search. Results are snapshots based on the search date. Does not perform deep code analysis or integration testing of the candidates.

    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 2 months ago

    Requires

    GitHub
    macOS

    Permissions

    Network Access
    Browser
    Read Files

    Allowed Hosts

    api.github.com
    Listed2 months ago

    What's inside

    Frequently Asked Questions

    Unlike static LLM responses, this skill forces the agent to fetch live repository metadata and commit history, ensuring recommendations are based on current maintenance status rather than outdated training data.

    Yes, you can define strict filters for license types (e.g., MIT, Apache 2.0), platform requirements like headless/CLI-only support, and specific architectural constraints like offline capability.

    The skill is designed for any AI agent with web search and HTTP request capabilities, as it relies on these to fetch real-time data from code repositories and technical documentation.

    You will receive the structured skill definition including the logic for multi-vector searching, metadata extraction protocols, and the ranking framework for comparison tables.

    The skill identifies 'not for' scenarios for each tool, highlighting specific use cases or environments where a particular piece of software is likely to fail or cause architectural friction.

    New in Research & Analysis