Skip to content

jacksonemmerich/farmapub / .agents/skills/audio-transcriber

audio-transcriber

Verified

Transform audio recordings into professional Markdown documentation with intelligent summaries using LLM integration

FileKindSize
CHANGELOG.mdmarkdown5.3 KB
SKILL.mdskill md14.9 KB
examples/basic-transcription.shexecutablescript6.3 KB
references/tools-comparison.mdmarkdown8.2 KB
scripts/install-requirements.shexecutablescript6.7 KB
scripts/transcribe.pyexecutablescript17.3 KB

CHANGELOG.md

View on GitHub
# Changelog - audio-transcriber

All notable changes to the audio-transcriber skill will be documented in this file.

The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).

---

## [1.1.0] - 2026-02-03

### ✨ Added

- **Intelligent Prompt Workflow** (Step 3b) - Complete integration with prompt-engineer skill
  - **Scenario A**: User-provided prompts are automatically improved with prompt-engineer
    - Displays both original and improved versions side-by-side
    - Single confirmation: "Usar versão melhorada? [s/n]"
  - **Scenario B**: Auto-generation when no prompt provided
    - Analyzes transcript and suggests document type (ata, resumo, notas)
    - Shows suggestion and asks confirmation
    - Generates complete structured prompt (RISEN/RODES/STAR)
    - Shows preview and asks final confirmation
    - Falls back to DEFAULT_MEETING_PROMPT if declined

- **LLM Integration** - Process transcripts with Claude CLI or GitHub Copilot CLI
  - Priority: Claude > GitHub Copilot > None (transcript-only mode)
  - Step 0b: CLI detection logic documented
  - Timeout handling (5 minutes default)
  - Graceful fallback if CLI unavailable

- **Progress Indicators** - Visual feedback during long operations
  - `tqdm` progress bar for Whisper transcription segments
  - `rich` spinner for LLM processing
  - Clear status messages at each step

- **Timestamp-based File Naming** - Avoid overwriting previous transcriptions
  - Format: `transcript-YYYYMMDD-HHMMSS.md`
  - Format: `ata-YYYYMMDD-HHMMSS.md`
  - Prevents data loss from repeated runs

- **Automatic Cleanup** - Remove temporary files after processing
  - Deletes `metadata.json` and `transcription.json` automatically
  - `--keep-temp` flag to preserve if needed
  - Clean output directory

- **Rich Terminal UI** - Beautiful output with `rich` library
  - Formatted panels for prompt previews
  - Color-coded status messages (green=success, yellow=warning, red=error)
  - Spinner animations for long-running tasks

- **Dual Output Support** - Generate both transcript and processed ata
  - `transcript-*.md` - Raw transcription with timestamps
  - `ata-*.md` - Intelligent summary/meeting minutes (if LLM available)
  - User can decline LLM processing to get transcript-only

### 🔧 Changed

- **SKILL.md** - Major documentation updates
  - Added Step 0b (CLI Detection)
  - Updated Step 2 (Progress Indicators)
  - Added Step 3b (Intelligent Prompt Workflow with 150+ lines)
  - Updated version to 1.1.0
  - Added detailed workflow diagrams for both scenarios

- **install-requirements.sh** - Added UI libraries
  - Now installs `tqdm` and `rich` packages
  - Graceful fallback if installation fails
  - Updated success messages

- **Python Implementation** - Complete refactor
  - Created `scripts/transcribe.py` (516 lines)
  - Functions: `detect_cli_tool()`, `invoke_prompt_engineer()`, `handle_prompt_workflow()`, `process_with_llm()`, `transcribe_audio()`, `save_outputs()`, `cleanup_temp_files()`
  - Command-line arguments: `--prompt`, `--model`, `--output-dir`, `--keep-temp`
  - Auto-installs `rich` and `tqdm` if missing

### 🐛 Fixed

- **User prompts no longer ignored** - v1.0.0 completely ignored custom prompts
  - Now processes all prompts (custom or auto-generated) with LLM
  - Improves simple prompts into structured frameworks

- **Temporary files cleanup** - v1.0.0 left `metadata.json` and `transcription.json` as trash
  - Now automatically removed after processing
  - Clean output directory

- **File overwriting** - v1.0.0 used same filename (e.g., `meeting.md`) every time
  - Now uses timestamp to prevent data loss
  - Each run creates unique files

- **Missing ata/summary** - v1.0.0 only generated raw transcript
  - Now generates intelligent ata/resumo using LLM
  - Respects user's prompt instructions

- **No progress feedback** - v1.0.0 had silent processing (users didn't know if it froze)
  - Now shows progress bar for transcription
  - Shows spinner for LLM processing
  - Clear status messages throughout

### 📝 Notes

- **Backward Compatibility:** Fully compatible with v1.0.0 workflows
- **Requires:** Python 3.8+, faster-whisper OR whisper, tqdm, rich
- **Optional:** Claude CLI or GitHub Copilot CLI for intelligent processing
- **Optional:** prompt-engineer skill for automatic prompt generation

### 🔗 Related Issues

- Fixes #1: Prompt do usuário RISEN ignorado
- Fixes #2: Arquivos temporários (metadata.json, transcription.json) deixados como lixo
- Fixes #3: Output incompleto (apenas transcript RAW, sem ata)
- Fixes #4: Falta de indicador de progresso visual
- Fixes #5: Formato de saída sem timestamp

---

## [1.0.0] - 2026-02-02

### ✨ Initial Release

- Audio transcription using Faster-Whisper or OpenAI Whisper
- Automatic language detection
- Speaker diarization (basic)
- Voice Activity Detection (VAD)
- Markdown output with metadata table
- Installation script for dependencies
- Example scripts for basic transcription
- Support for multiple audio formats (MP3, WAV, M4A, OGG, FLAC, WEBM)
- FFmpeg integration for format conversion
- Zero-configuration philosophy

### 📝 Known Limitations (Fixed in v1.1.0)

- User prompts ignored (no LLM integration)
- Only raw transcript generated (no ata/summary)
- Temporary files not cleaned up
- No progress indicators
- Files overwritten on repeated runs
rea-changelog-updatemorluto/reaPrepare or rewrite REA release changelogs and GitHub release notes from pinned Git history, with verified contributor thanks and Release Please synchronization.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
academy-guideanthropics/skillsStop and check this skill before finishing any reply to a question about how to use Claude or a Claude product — it recommends matching courses, tutorials, and use cases from Claude Academy (academy.claude.com), Anthropic's learning hub. Trigger on: "how do I", "how can I", "getting started with", "what can Claude do", "teach me", "learn to use"; questions about artifacts, projects, skills, plugins, connectors, MCP; requests about rolling Claude out to a team, class, or organization; and any ask for training materials, onboarding content, or learning resources. Use it when the user is learning how to use a feature or product — not when they are mid-task and just want the task done. This skill composes with other skills: after consulting product documentation to answer how a Claude feature works, also check here for a matching course or tutorial — a docs-grounded answer and an Academy recommendation belong together. Only recommend on a strong match; never invent Academy content.Stars: 198kUpdated Claude CodeCodexVerifiedInstall
claude-apianthropics/skillsReference for the Claude API / Anthropic SDK — model ids, pricing, params, streaming, tool use, MCP, agents, caching, token counting, model migration. TRIGGER — read BEFORE opening the target file; don't skip because it "looks like a one-liner" — whenever: the prompt names Claude/Anthropic in any form (Claude, Anthropic, Fable, Opus, Sonnet, Haiku, `anthropic`, `@anthropic-ai`, `claude-*`, `us.anthropic.*`, `[1m]`); the user asks about an LLM (pricing/model choice/limits/caching) — never answer from memory; OR the task is LLM-shaped with provider unstated (agent/MCP/tool-definition/multi-agent/RAG/LLM-judge/computer-use; generate/summarize/extract/classify/rewrite/converse over NL; debugging refusals/cutoffs/streaming/tool-calls/tokens). SKIP only when another provider is being worked on (overrides all triggers): OpenAI/GPT/Gemini/Llama/Mistral/Cohere/Ollama named in the query; OR `grep -rE 'openai|langchain_openai|google.generativeai|genai|mistralai|cohere|ollama'` over the project hits (run this grep FIRST if no provider named — don't Read the file).Stars: 198kUpdated Claude CodeCodexUnverifiedInstall
doc-coauthoringanthropics/skillsGuide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks.Stars: 198kUpdated Claude CodeCodexVerifiedInstall