agentleFS
fltmanfltman/claude-code-demo-creator/skills/elevenlabs-dialogue/SKILL.md

elevenlabs-dialogue

Skill19 starsChanged 5 months agoMITView on GitHub

Generate natural voice-over / narration audio with the ElevenLabs Text-to-Dialogue API. Use when producing a narrated demo video, podcast, audiobook, or any per-line voice-over where each line becomes its own mp3 and identical lines are cached (no re-billing). Sounds markedly better than plain text-to-speech. Reads ELEVENLABS_API_KEY from the environment.

  • Reads credentials
  • Installs packages

2 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions2 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 2

    1. AngelescuCiprian/claude-skills★ 0same name
    2. nikunj812/claude-skill★ 0same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If you narrate with someone's cloned voice, the honest default is to say so (e.g.

    line 79

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

ElevenLabs Dialogue

Turn a list of lines (each with a voice_id) into spoken audio using ElevenLabs' Text-to-Dialogue API — which produces more natural, expressive delivery than the plain /v1/text-to-speech endpoint, even for single-voice narration.

When to use

  • Voice-over for a demo video (one mp3 per shot, timed to each slide).
  • Multi-voice dialogue (podcasts, scenes) — each line carries its own voice.
  • Any narration where you want a per-line, content-addressed cache so re-runs are free for unchanged lines.

Prerequisites

pip install elevenlabs pydub        # pydub optional (cleaner concat); ffmpeg used otherwise
export ELEVENLABS_API_KEY=...        # never hardcode this in a committed file

The key lives in your environment. Find a cloned voice id with:

curl -s -H "xi-api-key: $ELEVENLABS_API_KEY" https://api.elevenlabs.io/v1/voices \
  | python3 -c 'import sys,json;[print(v["voice_id"],v["name"],v.get("category")) for v in json.load(sys.stdin)["voices"]]'

Generate

Input is a JSON array of { "text", "voice_id" } items. Keep [emotion] delivery tags (e.g. [warm], [calm]) inline — they're English even for other languages.

# one mp3 for the whole array:
python3 scripts/generate_dialogue.py --input lines.json --output out.mp3

# for a demo video, generate ONE file per shot so each slide is timed to its clip:
python3 scripts/generate_dialogue.py --input shot-01.json --output audio/01.mp3

lines.json:

[
  { "text": "[warm] Welcome — here's how it works.", "voice_id": "VOICE_ID" }
]

Useful flags: --list-chunks (print the chunking plan, no API calls), --regen 3,7 (force-regenerate specific chunks), --force (drop the whole cache), --no-concat (chunks only). The per-chunk cache lives in <output_dir>/.segments/<basename>/ — identical chunks are reused across runs, so you're never billed twice for the same line.

Verify (optional, recommended)

verify_dialogue.py transcribes each generated chunk back with whisper.cpp and flags TTS glitches — stuck loops, skipped/over-long output, dead air — so you catch a bad take before assembling the video.

WHISPER_CLI=~/whisper.cpp/build/bin/whisper-cli \
WHISPER_MODEL=~/whisper.cpp/models/ggml-base.bin \
WHISPER_LANG=en \
python3 scripts/verify_dialogue.py --cache-dir audio/.segments/01 --save-transcripts

For Swedish narration, point WHISPER_MODEL at a KB-Whisper ggml model and set WHISPER_LANG=sv for the most accurate check. Verification is optional — skip it if you don't have whisper.cpp installed.

Disclosure

If you narrate with someone's cloned voice, the honest default is to say so (e.g. open with a short "this is an AI using my cloned voice"). Use a clone only with that person's consent.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.