AI Tools

MCP server

To let an assistant run Cleanvoice jobs for you, connect the Cleanvoice MCP server at https://api.cleanvoice.ai/mcp. Every client, from Claude and ChatGPT to Claude Code, Codex, Cursor and VS Code, signs in with your Cleanvoice account.

See MCP server for setup per client, the tool list and limits.

Docs for AI coding assistants

Cleanvoice ships two machine-readable docs surfaces for AI tools:

• llms.txt — a compact curated summary of the API surface, common config options, and the recommended SDK-first workflow
• llms-full.txt — the full concatenated docs corpus for assistants that can handle more context

Start with llms.txt when you want a small prompt footprint. Use llms-full.txt when you want the broader docs corpus, more examples, or page-level wording.

https://docs.cleanvoice.ai/llms.txt
https://docs.cleanvoice.ai/llms-full.txt

Add to Claude Code

Append to CLAUDE.md in your project root:

curl -s https://docs.cleanvoice.ai/llms.txt >> CLAUDE.md

Claude will now have the Cleanvoice summary in project context. If you want the broader docs corpus instead, append https://docs.cleanvoice.ai/llms-full.txt.

Add to Cursor

mkdir -p .cursor/rules
curl -s https://docs.cleanvoice.ai/llms.txt > .cursor/rules/cleanvoice.mdc

Add to GitHub Copilot

curl -s https://docs.cleanvoice.ai/llms.txt >> .github/copilot-instructions.md

Any AI chat

Paste the URL into any model that supports URL fetching, or copy the raw contents directly into your chat context. Prefer llms.txt for concise context and llms-full.txt for deeper retrieval.

What gets loaded

• API endpoints and request/response patterns
• Common config options with their API, Python, and JavaScript names
• The recommended SDK-first submit -> poll -> download flow
• Multi-track and async workflow guidance
• Supported languages