11labs-acset

Wraps the ElevenLabs voice synthesis API as an ACSet (Attributed C-Set) schema, with Unison algebraic effect handlers for typed voice operations.

Use When

• Integrating ElevenLabs TTS into a category-theoretic data pipeline
• Building typed voice synthesis workflows with compile-time morphism checking
• Managing multi-voice narration with GF(3) trit conservation across voices
• Accessing ElevenLabs via MCP server (uvx elevenlabs-mcp)

Workflow

1. Set API key: export ELEVENLABS_API_KEY="sk_..."
2. Start MCP server: uvx elevenlabs-mcp (or python3 unified_elevenlabs_mcp_server.py)
3. Define voice triad: assign trit roles to voices (+1 generator, 0 coordinator, -1 validator)
4. Synthesize: call synthesize with text and voice, receiving typed Audio output
5. Verify conservation: confirm sum(trits) === 0 (mod 3) across voice set