Seed Audio 1.0 Skill

Generate speech, dialogue, ambience, sound effects, and mixed audio scenes with Seed Audio 1.0 through EvoLink.

Install

Primary Agent install:

skills.sh add evolink-seed-audio@latest

If skills.sh is unavailable, use the direct npm fallback:

npx evolink-seed-audio@latest -y --path ~/.claude/skills

Suppress browser opening while still printing the tracked key URL:

npx evolink-seed-audio@latest -y --no-open --path ~/.claude/skills

Agent Handoff

After installation, the installer prints:

EVOLINK_KEY_URL=<tracked key URL printed by the installer>
AGENT_NEXT_ACTION=open_key_url_then_collect_key

Open or show the printed EVOLINK_KEY_URL, ask the user to create or select an EvoLink key, save the pasted key as EVOLINK_API_KEY, validate it through the non-generating credits/account endpoint, then offer a starter prompt. Runtime installer key URLs use utm_source=skill&utm_medium=install; npm package-page key acquisition links use utm_source=npm&utm_medium=package.

Get an API Key

Open EvoLink API Keys, sign in or create an account, then create or copy an API key.

export EVOLINK_API_KEY="your_key_here"

Validate the key before generation:

scripts/seed-audio-generate.sh --validate-key

Starter Templates

Each starter template is designed to stay under 10 credits. Verify the live usage object and dashboard balance after the first real call.

| Template | Estimated credits | Prompt | |---|---:|---| | Starter template 1 | under 10 credits | Create a 10-second soft product audio bed with one calm narration line. | | Starter template 2 | under 10 credits | Create an 8-second clean app notification sound with a short spoken confirmation. |

Quick API Request

curl --request POST \
  --url https://api.evolink.ai/v1/audios/generations \
  --header "Authorization: Bearer ${EVOLINK_API_KEY}" \
  --header "Content-Type: application/json" \
  --header "X-EvoLink-Source: skill" \
  --header "X-EvoLink-Skill: seed-audio-1-0" \
  --header "X-EvoLink-Package: evolink-seed-audio" \
  --header "X-EvoLink-Campaign: seed-audio-1-0-skill" \
  --header "X-EvoLink-Touchpoint: first-run" \
  --data '{
    "model": "doubao-seed-audio-1-0",
    "prompt": "Create a 10-second soft product audio bed with one calm narration line.",
    "format": "mp3",
    "sample_rate": 24000
  }'

The API is asynchronous. The create request returns a task id; poll GET /v1/tasks/{task_id} until the task is completed, failed, or cancelled.

Commands

npx evolink-seed-audio@latest --llms
npx evolink-seed-audio@latest --skill
scripts/seed-audio-generate.sh --prompt "Short audio test" --dry-run

Links

• Create an EvoLink API key
• Try Seed Audio on EvoLink
• Official API docs

MIT. Repository: https://github.com/Evolink-AI/seed-audio-1-0-skill