transcribe

transcribe is a skill for Claude Code, Codex from greatSumini/cc-system. It costs 77 tokens per session (1,737 once invoked), scanned C, original, MIT.

A workflow for converting local audio recordings into Korean text using a speech-recognition model. It supports common audio formats, normalizes the sound first, and can use names and technical terms from the conversation as context.

In plain words
What is it for?
Transcribing Korean meetings, interviews, voice notes, and other local audio files such as M4A, MP3, WAV, or MP4.
Why use it?
It turns recordings into more readable Korean transcripts and reduces errors caused by uneven volume or quiet speakers. Context helps the recognizer handle important names and terms.

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/greatsumini/cc-system/transcribe
Any agent
npx skills add greatSumini/cc-system --skill transcribe
Clone the repo
git clone --depth 1 https://github.com/greatSumini/cc-system

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 transcribe

README.md
[![agentmods](https://agentmods.dev/badge/skills/greatsumini/cc-system/transcribe.svg)](https://agentmods.dev/skills/greatsumini/cc-system/transcribe)
Your own site
<a href="https://agentmods.dev/skills/greatsumini/cc-system/transcribe"><img src="https://agentmods.dev/badge/skills/greatsumini/cc-system/transcribe.svg" alt="Measured on agentmods" height="20"></a>
Per session 77 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,737 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 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 5.1 $0.00077 $0.01737
Opus 5 $0.00039 $0.00869
Sonnet 5 $0.00015 $0.00347
Haiku 4.5 $0.00008 $0.00174

Measured 6d ago against content hash 96a5b6ac82d2, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade C, and why

transcribe scanned grade C 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 6d 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.

Recursive force deletehighDestructive command

rm -rf with a variable or a broad path is one typo away from removing the wrong tree.

rm -rf "$OUTDIR" "$NORMALIZED"
.claude/skills/transcribe/SKILL.md · 129 lines

How it starts

The opening of the file, as written. The whole thing — 129 lines — stays where its author put it; the contents beside it link to each section on GitHub.

transcribe — 로컬 오디오 전사

Apple Silicon 네이티브 mlx-whisper(Whisper large-v3-turbo)로 로컬 오디오 파일을 한국어 텍스트로 전사한다.

플로우

1. 파일 경로 확인

사용자가 제공한 오디오 파일 경로를 확인한다. 상대경로면 현재 작업 디렉토리 기준 절대경로로 변환한다.

지원 포맷: m4a 기본, mp3/wav/mp4 등 ffmpeg이 디코딩 가능한 포맷 모두 동작.

2. 도메인 힌트 구성

현재 대화 맥락에서 다음 카테고리를 적극적으로 끌어모아 한국어 한두 문장으로 조합해 --initial-prompt에 넣는다. 맥락이 있으면 가능한 한 풍부하게 채운다(없는 정보를 지어내지는 않는다).

수집할 카테고리:

  • 참석자/화자 이름 (가능하면 전부)
  • 회사명·제품명·서비스명
  • 자주 쓰일 영문 약어·전문용어의 한글 표기 예시 (예: "토스페이먼츠를 'TPay'로도 부른다", "API를 '에이피아이'로 발음한다")
  • 도메인 키워드 5~20개

예시:

  • 대화에서 "최수민, 바이브마피아클럽, 구글, B2B 제안서"가 언급되었다면
    • --initial-prompt "이 녹음은 바이브마피아클럽 최수민과 구글의 B2B 제안 회의다. 등장 용어: 바이브마피아클럽(VMC), 구글, B2B 제안서, 견적, 라이선스."

맥락이 전혀 없으면 이 단계는 건너뛴다. 단, 사용자에게 "전사 품질을 올리려면 등장 인물·고유명사를 알려주세요"라고 한 번 권유한다.

3. 오디오 전처리

작은 목소리 화자 누락과 볼륨 편차로 인한 환각을 줄이기 위해 ffmpeg loudnorm으로 정규화한 뒤 전사한다. 이 단계는 항상 수행한다(추가 비용 거의 없음, 효과 큼).

NORMALIZED="$(mktemp -t transcribe_norm).wav"
ffmpeg -y -i "<AUDIO_PATH>" \
  -af loudnorm=I=-16:TP=-1.5:LRA=11 \
  -ar 16000 -ac 1 \
  "$NORMALIZED"
  • -ar 16000 -ac 1: Whisper 내부 표현(16kHz 모노)에 맞춰 미리 변환. 모델 입력 변환 비용 절감.
  • 사용자가 "잡음이 심하다", "BGM 깔려 있다"고 명시한 경우에만 추가로 demucs(음성 분리) 또는 arnndn(RNNoise) 단계를 적용한다. 기본 플로우에는 넣지 않는다 — 모델 다운로드/실행 비용이 크고 깨끗한 녹음에선 오히려 음성을 깎는다.

4. 전사 실행

임시 출력 디렉토리에 txt 포맷으로 내보낸다. 환각 방지 플래그를 기본 적용한다.

OUTDIR="$(mktemp -d -t transcribe)"
ORIG_BASE="$(basename "<AUDIO_PATH>")"; ORIG_BASE="${ORIG_BASE%.*}"
mlx_whisper \
  --model mlx-community/whisper-large-v3-turbo \
  --language ko \
  --output-format txt \
  --output-dir "$OUTDIR" \
  --output-name "$ORIG_BASE" \
  --condition-on-previous-text False \
  --temperature 0 \
  --no-speech-threshold 0.6 \
  --verbose False \
  [--initial-prompt "<DOMAIN_HINT>"] \
  "$NORMALIZED"

환각 방지 플래그 의도:

  • --condition-on-previous-text False: 한 번 잘못 인식한 텍스트가 뒤 구간으로 전염되는 문제 차단. 회의 녹음에선 거의 항상 이득.
  • --temperature 0: 샘플링 무작위성 제거. 동일 입력에 동일 출력 보장.
  • --no-speech-threshold 0.6: 무음/BGM 구간에서 "구독과 좋아요 부탁드립니다" 같은 환각 생성 방지. 기본값(0.6)을 명시적으로 박아둔다.

Read the full file on GitHub · 129 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. 6d ago First seen · 129 lines · 77 tokens per session scan C 96a5b6ac82d2

Subscribe to this mod's changes

transcribe is a skill published in the GitHub repository greatSumini/cc-system (436 stars, last pushed 2mo ago), licensed MIT. It adds 77 tokens to every session and 1,737 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

local-ai-agents

Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…

microsoft/ai-agents-for-beginners · 200 tokens

next-cache-components-adoption

Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…

vercel/next.js · 95 tokens

next-cache-components-optimizer

Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…

vercel/next.js · 170 tokens

next-partial-prefetching-adoption

Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…

vercel/next.js · 103 tokens

chronicle

Analyze Copilot session history for standup reports, usage tips, session search, and session reindexing. Use when the user asks for a standup, daily summary, usage tips, workflow recommendations, wants to search or find past sessions by keyword/file/PR, wants to reindex their session store, or asks about deleting…

microsoft/vscode · 72 tokens