Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Yap Transcribe Recordings

    1

    Turn voice memos, meetings, interviews and lectures into transcripts, notes and action items, privately on your computer.

    Free

    1 useSecurity scanned
    Yap Transcribe Recordings

    Yap Transcribe Recordings

    Example session with this skill installed

    Transcribe ~/Desktop/team-meeting.m4a and list the action items with owners.

    • Read your context and instructions
    • Compiled the yap transcribe recordings

    Summary

    The weekly product sync covered the launch checklist, a pricing question and the support backlog. Transcribed locally with Yap MCP (language detected: English, 41 minutes, nothing uploaded).

    Action items

    OwnerTaskDue
    MayaFinalize the launch checklist and share it with the teamFriday
    JonDraft the pricing FAQ from the questions raised in the meetingNext Tuesday
    SamClear the five oldest support ticketsThursday

    Decisions

    • Launch date stays as planned.
    • Pricing page copy will be reviewed once the FAQ is ready.

    Transcript excerpt

    [00:00:12] Maya: Let us start with the launch checklist, we are two items away from done.
    [00:03:40] Jon: On pricing, customers keep asking what happens after the free start.
    [00:09:05] Sam: The support backlog is down to nine tickets, five are over a week old.

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

    About this skill

    Transcribing meetings or voice memos often requires uploading sensitive data to the cloud, risking privacy and data security. This skill utilizes the Yap MCP server to process audio files locally on your machine. It handles everything from raw transcription to structured meeting notes without your data ever leaving your local environment.

    What it does

    • Local transcription converts audio and video files into text using NVIDIA Parakeet models.
    • Automated summarization creates condensed versions of long recordings, organized by timestamped sections.
    • Action item extraction identifies decisions, owners, and deadlines directly from the spoken text.
    • Dictation search queries the history of the Yap dictation app to find past recordings by keyword or date.
    • Multilingual support automatically detects and transcribes content in 25 European languages.

    How it works

    1. Provide the absolute path to an audio or video file located on your local drive.
    2. The skill invokes the transcribe_audio tool to generate a text output with optional timestamps.
    3. For existing dictations, use the search_dictations command with a keyword query and a day-range window.
    4. The output is formatted according to your request: clean transcripts, summaries, or structured notes.

    Frameworks & tools

    This skill requires the Yap MCP server (npx -y github:founden-bot/yap-mcp) and utilizes NVIDIA Parakeet for local speech-to-text processing.

    Why this beats prompting it yourself

    Generic prompts cannot access local file paths or interface with the Yap MCP tools to perform the actual heavy lifting of audio processing. This skill bridges the gap between your LLM and your local filesystem, ensuring zero-latency data transfers and total privacy.

    Use cases

    • Transcribe long interviews with accurate timestamps for easy reference.
    • Generate structured meeting minutes with clear owners and due dates from a Zoom recording.
    • Locate specific past voice memos by searching through dictation history.
    • Convert multi-language lecture recordings into localized summaries.

    Known limitations

    The first run requires a one-time 670 MB model download. It is limited to the 25 European languages supported by the local Parakeet model.

    Requires Yap MCP server installation. Speech models require a one-time 670 MB download. Limited to 25 European languages.

    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

    1 use

    Times used to date

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 29 days ago

    • Free with an Agensi account

    Requires

    Yap MCP
    Node.js 20+
    github
    Node.js

    Permissions

    Network Access
    Listed29 days ago

    What's inside

    Frequently Asked Questions

    This skill uses the Yap MCP server to process audio locally on your machine, ensuring your sensitive meeting data and voice memos never leave your environment or get uploaded to the cloud.

    The skill includes tools for local transcription via NVIDIA Parakeet, automated meeting summarization, action item extraction, and a specialized search function for your Yap dictation history.

    It supports 25 European languages, including English, Spanish, French, German, and Italian, with automatic language detection during the transcription process.

    You will need to have Node.js installed to run the Yap MCP server (`npx -y github:founden-bot/yap-mcp`) and allow for a one-time 670 MB model download during the initial setup.

    Yes, because it functions as an MCP (Model Context Protocol) tool, it is compatible with any AI agent or LLM interface that supports the MCP standard, such as Claude Desktop.

    Unlike manual prompting, this skill provides the necessary bridge to your local filesystem, allowing the AI to actually 'hear' and process local files that a standard prompt cannot access.

    New in LLM Ops