# Voice Message Transcriber

> Detects new local voice notes or audio files, triggers Whisper transcriptions, and prepares structured text outputs for conversational agent consumption.

- Skill: `knuckles-team/voice-message-transcriber` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add knuckles-team/voice-message-transcriber`
- Raw SKILL.md: https://api.skillmd.com/api/skills/knuckles-team/voice-message-transcriber/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: Knuckles-Team (https://skillmd.com/u/knuckles-team)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/knuckles-team/voice-message-transcriber

---


# Voice Message Transcriber Workflow

**CONCEPT:KG-2.12**

Detects new local voice notes or audio files, triggers Whisper transcriptions, and prepares structured text outputs for conversational agent consumption.

## Steps

### Step 0: Audio Transcriber Mcp
**Agent**: `intake-agent`
**Tools**: `graph_query, nc_files`

Transcribe an audio file or capture live microphone input using the transcribe_audio tool.
Expected: `raw_transcript`

### Step 1: User Interaction
**Agent**: `processor-agent`
**Tools**: `graph_analyze, document_tools`

Present the raw transcript to the user. Ask if they want to translate it, format it, or route it directly to another specialist agent for text response.
Expected: `structured_transcript, target_agent`

### Step 2: KG Persistence [depends_on: user-interaction]
**Agent**: `processor-agent`
**Tools**: `graph_write`

Persist workflow results as nodes and edges in the Knowledge Graph.
Create appropriate typed nodes with metadata and link to existing domain entities.

## Output
- Voice Message Transcriber results persisted in KG
- Structured report (MD/PDF)
- Audit trail with timestamps and agent attributions

## Human Oversight Required
✅ Critical decisions require human review and approval.

## Execution

Run this workflow as a dependency-ordered DAG. Steps with no unmet `depends_on` run in parallel; dependents run after their prerequisites complete.

- **Run first (in parallel):** Step 0 — Audio Transcriber Mcp; Step 1 — User Interaction
- **After level 0:** Step 2 — KG Persistence

**Execution:** If graph-os is reachable, offload the whole DAG via `graph_orchestrate action=execute_workflow` (or the `kg-delegate` skill) for true parallel/swarm execution. Otherwise execute the steps natively in dependency order: run steps with no unmet `depends_on` in parallel, then their dependents.