ingest-audio

ingest-audio is a skill for Claude Code, Codex from stevepridemore/graph-memory. It costs 56 tokens per session (1,237 once invoked), scanned A, original, MIT.

A workflow that transcribes a local audio or video file with Whisper, a speech-to-text tool, and adds the result to a memory graph.

In plain words
What is it for?
Use it with MP3, WAV, M4A, MP4, and similar files, choosing a Whisper model and adding topic or author details.
Why use it?
It removes the need to listen through recordings and manually copy their contents into your knowledge system.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/stevepridemore/graph-memory/ingest-audio
Any agent
npx skills add stevepridemore/graph-memory --skill ingest-audio
Clone the repo
git clone --depth 1 https://github.com/stevepridemore/graph-memory

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for ingest-audio

README.md
[![agentmods](https://agentmods.dev/badge/skills/stevepridemore/graph-memory/ingest-audio.svg)](https://agentmods.dev/skills/stevepridemore/graph-memory/ingest-audio)
Your own site
<a href="https://agentmods.dev/skills/stevepridemore/graph-memory/ingest-audio"><img src="https://agentmods.dev/badge/skills/stevepridemore/graph-memory/ingest-audio.svg" alt="Measured on agentmods" height="20"></a>
Per session 56 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,237 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00056 $0.01237
Opus 5 $0.00028 $0.00619
Sonnet 5 $0.00011 $0.00247
Haiku 4.5 $0.00006 $0.00124

Measured 4d ago against content hash 2e6867d3aafb, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

ingest-audio scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/ingest-audio/SKILL.md · 128 lines

How it starts

The opening of the file, as written. The whole thing — 128 lines — stays where its author put it; the contents beside it link to each section on GitHub.

The user wants to transcribe a local audio or video file and ingest it into the graph memory system.

Arguments: $ARGUMENTS

Step 1: Parse arguments

  • First positional argument: local file path (required)
  • --model: Whisper model size. Default: base. Options: tiny, base, small, medium, large
    • tiny: fastest, least accurate (~39MB)
    • base: good balance, recommended default (~74MB)
    • small: noticeably better accuracy (~244MB)
    • medium: high accuracy, slow on CPU (~769MB)
    • large: best accuracy, very slow on CPU (~1.5GB)
  • --topic: topic hints for metadata (comma-separated)
  • --author: speaker or creator name
  • --now: process immediately inline after transcription instead of queuing

If no file path is provided, ask the user for one.

Step 2: Verify Whisper is installed

Run: whisper --help

If this fails, report:

Whisper is not installed. Install it with:
  pip install openai-whisper

Note: This downloads model weights (~74MB for 'base') on first run.
ffmpeg is also required:
  - Windows: winget install ffmpeg
  - macOS:   brew install ffmpeg
  - Linux:   apt install ffmpeg (or your distro's equivalent)

Then stop.

Step 3: Verify the file exists and is a supported format

Supported: .mp3, .wav, .m4a, .mp4, .ogg, .flac, .webm, .mkv, .avi, .mov

If the file doesn't exist or the format isn't supported, report the issue and stop.

Step 4: Transcribe with Whisper

Output directory: ~/graph-memory/.tmp/graph-audio-ingest/

Run:

whisper "<file-path>" --model <model> --output_format txt --output_dir "~/graph-memory/.tmp/graph-audio-ingest/"

Whisper outputs <original-filename>.txt in the output directory.

Note: This may take several minutes for longer files on CPU. Inform the user that transcription is running.

If transcription fails, report the error and stop.

Step 5: Read and validate the transcript

Read the output .txt file. If it's empty or very short (under 20 words), warn the user:

  • "Transcript appears empty or very short. The audio may be silent, too quiet, or in a different language."
  • Suggest trying --model small or --model medium for better accuracy.

Read the full file on GitHub · 128 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 4d ago First seen · 128 lines · 56 tokens per session scan A 2e6867d3aafb

Subscribe to this mod's changes

ingest-audio is a skill published in the GitHub repository stevepridemore/graph-memory (2 stars, last pushed 3mo ago), licensed MIT. It adds 56 tokens to every session and 1,237 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.