rec-subtitle

A macOS command that turns speech in a video into an SRT subtitle file. SRT is a common timed-caption format; subtitles can also be permanently drawn onto the video.

In plain words
What is it for?
Use it to create caption files for YouTube, permanently burn captions into videos, and improve recognition of proper nouns with a vocabulary file.
Why use it?
It saves you from transcribing and timing captions by hand. You can provide optional edited captions or transcribe a microphone-only track for cleaner results.

Command

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 commands/esaruoho/apple/rec-subtitle
Clone the repo
git clone --depth 1 https://github.com/esaruoho/apple
Per session 73 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 606 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. 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 $0.00073 $0.00606
Opus 5 $0.00036 $0.00303
Sonnet 5 $0.00015 $0.00121
Haiku 4.5 $0.00007 $0.00061

Measured 3d ago against content hash 119d314bffcf, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

rec-subtitle scanned grade A with 0 findings 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 3d 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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

commands/rec-subtitle.md · 35 lines

What it actually says

Run the apple-skill rec-subtitle tool on $ARGUMENTS.

Use Bash to execute (one call, then stop):

/Users/esaruoho/work/apple/bin/rec-subtitle $ARGUMENTS
  • /rec-subtitle <video> → transcribes the audio → <stem>.srt. Upload the video to YouTube and add this .srt as a caption track (soft subtitles — viewers can toggle them).
  • /rec-subtitle <video> --burn → transcribes if needed, then hard-burns the subtitles into <stem>-subtitled.mov (white text, black outline, bottom-center; timed per cue). Use for platforms without a caption track, or when the captions must be permanent.
  • --srt FILE uses an existing/edited .srt instead of transcribing.
  • --mic <stem>-mic.m4a transcribes the voice-only track (from rec-audio split) for the cleanest transcript — best when the recording also has loud system audio.
  • --model NAME picks the Whisper model (default: whisp's own default).
  • Proper nouns are aligned automatically. bin/recburn-vocabulary.json (or ~/.config/recburn/vocabulary.json, or .recburn-vocabulary.json beside the recording) both biases Whisper (--initial_prompt) and sweeps the finished .srt, so Paketti / Renoise / Lackluster are never "Pucketty" / "Reno" / "Lacklustre". Unlisted mishearings are caught by sound (Soundex); real English words are never rewritten. --vocab FILE uses another vocabulary, --no-vocab turns it off.
  • --fix-srt <file.srt> aligns a transcript you already have, in place (no video needed).
  • --self-test checks the vocabulary rules headlessly (no media, no Whisper).

Transcription runs Whisper locally via ~/work/whisp/whisp; burn-in is Apple-native AVFoundation (video re-encoded). After the command completes, report only the lines it printed.

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. 3d ago First seen · 35 lines · 73 tokens per session scan A 119d314bffcf

Subscribe to this mod's changes

rec-subtitle is a command published in the GitHub repository esaruoho/apple (11 stars, last pushed 6d ago), licensed MIT. It adds 73 tokens to every session and 606 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.