Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    daedalus

    1

    Diagnose and configure audio notifications and speech feedback for Claude Code sessions.

    Free

    Security scanned
    daedalus

    daedalus

    Example session with this skill installed

    Daedalus isn't making any sounds when I'm in another window. Can you check why?

    • Read your context and instructions
    • Compiled the daedalus

    I will check the system status and recent logs to see why the audio notifications are failing during window switches.

    Diagnostic Report

    ComponentStatusDetails
    DaemonRunningPID: 48291, Port: 47113
    Audio BackendafplayResolved (macOS)
    Speech BackendsayResolved (macOS)
    Focus BackendosascriptResolved
    Config ErrorsNoneAll keys in .toml are valid

    Log Analysis (Last 5 events):

    • perm focus=focused label=- -> earcon=none speech=False (terminal focused; a tone is enough)
    • perm focus=unfocused label=daedalus -> earcon=needs_you speech=True (waiting on you, and you're away)
    • turn took 1.2s, under the 3.0s threshold (short turns are silent by design)

    The logs show that turns under 3.0 seconds are being suppressed by your min_turn_seconds setting. Additionally, when your terminal is focused, speech is disabled because the system assumes you are watching the output.

    Next steps

    1. Run daedalus test to confirm your speakers are working.
    2. Lower min_turn_seconds in ~/.daedalus.toml if you want sounds for short tasks.
    3. Check if mute was accidentally triggered by running daedalus unmute.

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

    About this skill

    The Daedalus skill helps you manage and fix the Daedalus audio plugin for Claude Code. Users often struggle with silent notifications, unwanted speech, or incorrect focus detection when using multiple terminal panes. This tool provides the specific commands and diagnostic steps needed to tune the background daemon and its audio backends.

    What it does

    • System diagnostics through a doctor command to identify missing audio, speech, or focus dependencies.
    • Log analysis to explain why specific sounds were triggered or suppressed based on turn duration and window focus.
    • Configuration management for adjusting turn duration thresholds, sound packs, and speech escalation.
    • State control to persistently mute or unmute audio feedback and test speaker output.

    How it works

    1. Run daedalus doctor to check the status of audio backends and daemon connectivity.
    2. Inspect ~/.daedalus/daedalus.log to find the specific reasoning behind recent audio decisions.
    3. Edit ~/.daedalus.toml to modify behavior like min_turn_seconds or sound packs.
    4. Use the mute or unmute commands to toggle the persistent state of the background monitor.

    Frameworks & tools

    This skill is purpose-built for Claude Code. It interacts with Python 3.11+, TOML configuration files, and OS-native audio utilities like afplay, paplay, spd-say, and xprop.

    Why this beats prompting it yourself

    Generic AI models do not know the internal logic of the Daedalus daemon or its specific log format. This skill correctly interprets complex log strings like focus labels and loopback socket states that standard prompts would hallucinate.

    Use cases

    • Fix missing sound notifications when the terminal is not in focus.
    • Stop the agent from speaking during short, repetitive turns.
    • Change the audio pack to a quieter volume for open-office environments.
    • Troubleshoot why speech-to-text isn't working on specific Linux distributions.

    Known limitations

    Only works with Claude Code. Requires local installation of the Daedalus plugin in the user home directory. Speech on Linux requires specific packages like speech-dispatcher or espeak.

    Works only with Claude Code. Requires Python 3.11+ and specific OS audio tools. Plugin must be installed in ~/.claude/skills/ to function.

    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 1 month ago

    • Free with an Agensi account

    Requires

    Python

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Network Access
    Environment Variables
    Listed1 month ago
    Updated23 days ago

    What's inside

    Frequently Asked Questions

    Daedalus is purpose-built to configure and troubleshoot the audio plugin for Claude Code, specifically addressing issues like silent notifications, unwanted speech during short turns, and incorrect terminal focus detection.

    This skill is exclusively designed for use with Claude Code and requires Python 3.11+ along with the Daedalus plugin installed in your home directory.

    The skill provides automated diagnostic commands like 'daedalus doctor' to verify your audio backends and helps you correctly modify the .toml configuration files and system-level audio dependencies.

    Yes, Daedalus supports native audio utilities across platforms, including 'afplay' for macOS and 'paplay' or 'spd-say' for Linux, though some Linux distributions may require additional packages like speech-dispatcher.

    Unlike generic prompts, this skill understands the specific log formats and socket states of the Daedalus daemon, allowing it to accurately interpret focus labels and loopback states without hallucination.

    New in Agents & Orchestration