Let an agent do the transcribing.

AFK gives a coding agent a local command, clean streams, and a small set of rules for handling files. Install the bundled Agent Skill explicitly, then let the agent call the CLI you already use.

Install the bundled skill

The signed AFK app includes a short afk-transcription Agent Skill. Install it from the Automation screen or from Terminal:

$ afk install-agent-skill

The command copies the skill to ~/.agents/skills/afk-transcription. An identical copy is left alone. A different existing copy is not overwritten unless you deliberately pass --force.

Download the bundled Agent Skill (SKILL.md) →

Give the agent a bounded workflow

  1. Pass one quoted local path to afk transcribe.
  2. Keep transcript output on stdout and diagnostics on stderr.
  3. Use --json or --srt when the next step needs structured output.
  4. Use a shell loop for a folder; AFK does not provide a batch flag or MCP server.
$ afk transcribe recording.m4a > recording.txt
$ afk transcribe lesson.mp4 --srt > lesson.srt

Respect the input boundary

The CLI accepts AFK's supported audio inputs and an .mp4 file with an audio track. It extracts MP4 audio into a temporary directory and removes it after the command. Other video containers are not supported. An agent should not promise a MOV import or send a transcript to a networked tool without an explicit instruction.

Keep local and network steps distinct

First-run model preparation may need the network. Once the model is ready, recognition and file reading run locally. A shell command or agent may choose to send stdout somewhere else; that is a separate action outside AFK's transcription path.

Agent handoff: the CLI guide owns flags, streams, and exit behavior. This page owns the skill and the workflow around it.

See the folder-loop use case →