Skip to content

Conversation

@triepod-ai
Copy link

Summary

This PR adds MCP tool annotations to all 9 tools in the MiniMax MCP server, improving AI assistant understanding of tool capabilities.

Changes

Added ToolAnnotations to each @mcp.tool() decorator with:

  • title: Human-readable tool names
  • readOnlyHint: Distinguishes read vs write operations
  • destructiveHint: All false (tools create content, don't delete)
  • openWorldHint: True for API-calling tools, false for local-only

Tools Annotated

Tool Title readOnlyHint openWorldHint
text_to_audio Text to Audio false true
list_voices List Voices true true
voice_clone Voice Clone false true
play_audio Play Audio true false
generate_video Generate Video false true
query_video_generation Query Video Generation true true
text_to_image Text to Image false true
music_generation Music Generation false true
voice_design Voice Design false true

Why These Annotations Matter

Per the MCP specification:

  • readOnlyHint: Helps AI assistants understand which tools only read data vs. which create content
  • destructiveHint: All set to false since these tools create new content rather than delete
  • openWorldHint: Indicates which tools make external API calls (important for cost awareness)

🤖 Generated with Claude Code

This PR adds MCP tool annotations to improve AI assistant understanding
of tool capabilities:

- title: Human-readable tool names
- readOnlyHint: true for list_voices, play_audio, query_video_generation
- readOnlyHint: false for generation tools (create new content)
- destructiveHint: false for all tools (creation, not destruction)
- openWorldHint: true for API-calling tools (external service)
- openWorldHint: false for play_audio (local only)

Tools annotated:
1. text_to_audio - Generate audio from text
2. list_voices - List available voices (readOnly)
3. voice_clone - Clone a voice from audio
4. play_audio - Play local audio file (readOnly, local only)
5. generate_video - Generate video from prompt
6. query_video_generation - Check video status (readOnly)
7. text_to_image - Generate image from prompt
8. music_generation - Generate music from prompt/lyrics
9. voice_design - Design a new voice

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants