ComfyUI-WhisperCPP AGENTS.md

Project instructions for a ComfyUI custom node that transcribes audio with whisper.cpp, with optional speech alignment and speaker identification. ComfyUI is a visual system for building media-processing workflows from connected nodes.

In plain words
What is it for?
Use them when modifying audio loading, transcription, model downloads, vocal separation, word alignment, diarization, or ComfyUI node registration.
Why use it?
They require changes to be tested and verified in a real workflow, reducing the risk of relying on guessed API details or untested behavior.

Instructions file for CodexOpenCode

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 instructions/obvirm/comfyui-whispercpp/agents-md
Clone the repo
git clone --depth 1 https://github.com/obvirm/ComfyUI-WhisperCPP

Made for: Codex, OpenCode.

Per session 2,028 This file is loaded in full into every session.
When invoked 2,028 The same file — it is already loaded in full.
Security scan A 1 finding. 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 $0.02028 $0.02028
Opus 5 $0.01014 $0.01014
Sonnet 5 $0.00406 $0.00406
Haiku 4.5 $0.00203 $0.00203

Measured 2d ago against content hash 217fdf12ad5d, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

ComfyUI-WhisperCPP AGENTS.md scanned grade A with 1 finding 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 2d 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.

Runs shell commandslowCapability

Expected in a hook, worth knowing in a rule or an instructions file.

grep -rn "os.system\|subprocess" --include="*.py"
AGENTS.md · 185 lines

How it starts

The opening of the file, as written. The whole thing — 185 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md — ComfyUI-WhisperCPP

Tujuan

Custom node ComfyUI untuk transkripsi audio pakai whisper.cpp (C API via ctypes), tanpa whisperx.
Full pipeline: ASR (whisper.cpp native) → Alignment opsional (torchaudio wav2vec2) → Diarization opsional (pyannote.audio).

Aturan Main

1. GAK BOLEH NGE-GAS TANPA VERIFIKASI

  • Setiap perubahan: test dulu. Jangan ngarep "mungkin work".
  • End-to-end: bikin workflow ComfyUI beneran, inject prompt, verifikasi output.
  • Jangan pernah nebak properti API tanpa riset dulu.

2. Struktur Folder

ComfyUI-WhisperCPP/
├── __init__.py              # register node, NODE_CLASS_MAPPINGS
├── whispercpp_node.py       # class WhisperCPPNode (ComfyUI node)
├── whispercpp/
│   ├── __init__.py
│   ├── whisper_lib.py       # ctypes binding ke whisper.cpp (C struct, load, transcribe)
│   ├── model.py             # download manager (GGML_MODEL_KEYS, progress bar)
│   └── audio.py             # audio loader / preprocessor
│   └── ext/
│       ├── __init__.py
│       ├── uvr.py          # vocal separation via audio-separator (UVR MDX-Net)
│       ├── alignment.py     # forced alignment via torchaudio wav2vec2 (opsional)
│       └── diarization.py   # speaker diarization via pyannote (opsional)
├── js/
│   └── whispercpp_node.js   # frontend toggle (show_advance_cpp, show_advance_ext)
├── whispercpp.json          # model config
└── MEMD.md                  # project memory

3. Output WAJIB lewat Node Connection

  • filename_prefix dan output_format DILARANG — output cuma lewat 7 socket: text, segments_json, srt, vtt, tsv, aud, json_result
  • Semua format generate di memory (_make_outputs), gak ada file I/O.

4. JS Toggle (Frontend)

  • Dua toggle independent: show_advance_cpp dan show_advance_ext.
  • Approach: grab widget reference di onNodeCreated, simpen di closure.
  • Show/Hide: JANGAN pakai widget.hidden (belum tentu real).
  • JANGAN main-main this.widgets dengan cara yang bikin reference ilang.
  • Selalu setTimeout biar widget kebentuk dulu.

Read the full file on GitHub · 185 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. 2d ago First seen · 185 lines · 2,028 tokens per session scan A 217fdf12ad5d

Subscribe to this mod's changes

ComfyUI-WhisperCPP AGENTS.md is an instructions file published in the GitHub repository obvirm/ComfyUI-WhisperCPP (10 stars, last pushed 24d ago), licensed Apache-2.0. It adds 2,028 tokens to every session, about $0.0101 per session on Opus 5. A static security scan graded it A with 1 finding (runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.