Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    classnotes pro

    1

    Automated lecture capture, local transcription, and exam synthesis for macOS.

    $20

    /mo

    Subscription · cancel anytime

    Secure checkout via Stripe

    • Always the latest version
    • Delivered live to your agent via MCP
    • Cancel anytime, access ends at period end
    Security scanned
    classnotes pro

    classnotes pro

    Example session with this skill installed

    I just finished the lecture for CS101. Here is the recording and the slides folder. Can you generate the lecture note and update my exam synthesis?

    • Read your context and instructions
    • Compiled the classnotes pro

    I have processed the recording for CS101 dated 2026-10-12. The transcription is complete, and I have cross-referenced the 12 captured slide frames. ## Lecture Note: Distributed Systems — CS101 **Date:** 2026-10-12 · **Lecturer:** Dr. Aris · **Source:** transcripts/2026-10-12.txt | Section | Details | | :--- | :--- | | **In one line** | Introduction to CAP theorem and its implications on distributed consistency. | | **Key concepts** | **Eventual Consistency:** A consistency model where all nodes eventually converge to the same value [from slide 4]. | | **Formulas** | **CAP Theorem:** $C + A + P \neq 3$. You can only pick two properties in a partitioned network [from slide 7]. | | **⚡ Exam signals** | "This trade-off is the heart of every midterm question on architecture." | | **Self-test** | 1. Why is P mandatory in a distributed system? [Bloom: Analysis] | **Exam Synthesis Updated** The `SYNTHESIS.md` for CS101 has been rebuilt. It now includes 8 lectures and flags "Partition Tolerance" as a high-density exam topic. **Next steps** - Review the **⚡ Exam signals** section in the new note. - Run `make-anki.py` to update your flashcard deck with the new concepts. - Check `ASSIGNMENTS.md` for the Lab 2 deadline mentioned at the end of the session.

    $20/mo

    Subscription · always the latest version

    • Secure checkout via Stripe
    • Cancel anytime
    • Delivered live via MCP

    About this skill

    University lectures are messy. Audio is muffled, slides move too fast, and crucial exam hints get lost in the noise. Taking manual notes while trying to listen is a losing battle that leaves you with gaps and no clear revision path.

    This skill automates the entire lecture-to-study-guide pipeline on macOS. It captures system audio and slides, generates local transcripts using whisper.cpp, and cross-references everything to build high-fidelity notes. Beyond simple summaries, it identifies "Exam Signals" where lecturers emphasize testable material and maintains a running Synthesis file that updates after every class. All processing happens locally on your machine to ensure privacy and offline availability.

    What it does

    • Local Transcription converts lecture audio to text using whisper.cpp without cloud uploads.
    • Visual Capture grabs and deduplicates slides from meeting windows or browser tabs.
    • Exam Signal Detection greps transcripts for specific phrases that indicate testable content.
    • Automated Synthesis rebuilds a master course guide from scratch after every new lecture.
    • Study Tool Generation produces Anki-ready flashcards, mock exam papers, and cheat sheets.
    • Deadline Tracking extracts assignments and dates into a centralized schedule.

    How it works

    1. Initialize the environment by running the install script to fetch models and set up the local directory structure.
    2. Capture live sessions by selecting the class window; the tool records audio and snaps slides in the background.
    3. Process recordings or raw transcripts to generate cleaned, timestamped text files.
    4. Verify outputs using automated gates that flag unsourced numbers, proper nouns, or missing slide references.
    5. Rebuild the course synthesis to provide an up-to-date exam prep file that links concepts across the whole term.

    Frameworks & tools

    Built for macOS environments utilizing whisper.cpp for local inference and ffmpeg for audio handling. It uses Python scripts for data processing, BlackHole for system audio routing, and shell scripts to orchestrate the pipeline.

    Why this beats prompting it yourself

    Generic AI summaries often hallucinate facts or miss subtle verbal cues about exams. This skill uses a strict verification gate against local transcripts and deduplicated slide frames to ensure every claim is grounded. It also manages the state of an entire semester, linking notes into a single synthesis file that manual prompting cannot track over time.

    Use cases

    • Live Class Capture for remote or hybrid lectures where you need to focus on listening.
    • Exam Prep for generating a comprehensive study guide from a full semester of notes.
    • Catching Up by identifying missing sessions and open questions from missed classes.
    • Active Recall through automated generation of Anki cards and mock exam questions.

    Known limitations

    Requires macOS. Live audio capture requires one-time manual setup of Multi-Output devices. Transcription speed depends on local CPU/GPU resources and lacks speaker diarization.

    Requires macOS. Live audio capture requires manual setup of Multi-Output devices. Transcription speed depends on local hardware. No speaker diarization.

    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

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 28 days ago

    Requires

    Groq
    Hugging Face
    macOS

    Permissions

    Terminal / Shell
    Network Access
    Read Files
    Write Files
    Environment Variables

    Allowed Hosts

    api.groq.com
    huggingface.co
    Listed28 days ago

    What's inside

    Frequently Asked Questions

    Classnotes Pro performs all audio transcription and data processing locally on your macOS device using whisper.cpp, ensuring your lecture recordings never leave your machine or touch the cloud.

    The skill requires macOS and utilizes BlackHole for audio routing; while the initial setup requires a one-time configuration of Multi-Output devices, the included install script automates the model fetching and directory setup.

    Unlike standard AI summaries, this skill uses a "Verification Gate" to cross-reference transcripts with captured slides and a "Synthesis" engine that tracks concepts across the entire semester rather than just a single session.

    You receive the full suite of automation scripts, the local transcription environment, the synthesis engine, and a set of templates for generating Anki cards and mock exams.

    Yes, the skill is designed to work with both live virtual lectures via browser or meeting apps and pre-recorded video files, provided they are played through your macOS system.

    The tool identifies "Exam Signals"—specific verbal cues and emphasized phrases from your lecturer—and automatically tags them so you can prioritize testable material during revision.

    New in Productivity