@avala-ai/mcp-server
v0.7.0
Published
MCP server for the Avala API
Readme
@avala-ai/mcp-server
Model Context Protocol (MCP) server for the Avala API. Lets AI assistants (Claude, etc.) interact with your Avala annotation data.
Installation
npm install -g @avala-ai/mcp-serverRequires Node.js 18+.
Setup
Set your API key:
export AVALA_API_KEY="avk_your_api_key"MCP is read-only unless you explicitly enable mutation tools:
export AVALA_MCP_ENABLE_MUTATIONS=trueClaude Desktop
Add to your claude_desktop_config.json:
{
"mcpServers": {
"avala": {
"command": "npx",
"args": ["-y", "@avala-ai/mcp-server"],
"env": {
"AVALA_API_KEY": "avk_your_api_key"
}
}
}
}Claude Code
claude mcp add avala -- npx -y @avala-ai/mcp-serverDirect Usage
avala-mcp-serverAvailable Tools
| Tool Category | Description |
|---------------|-------------|
| Datasets | List and inspect datasets, sequences, frames, calibration, and ingest health |
| Projects | List and inspect projects |
| Exports | List, inspect, and create annotation exports |
| Fleet | Device management, recordings, events, alerts, rules |
| Agents | List and inspect automation agents |
| Webhooks | List and inspect webhook subscriptions |
| Storage | List and test storage configurations |
| Quality | Quality targets and consensus scoring |
| Annotation Issues | List and manage annotation issue data |
| Organizations | List organizations and members |
| Slices | List and inspect data slices |
| Stats | Get overview statistics for your account |
| Workflows | Composite tools: fleet health overview, project quality summary, workspace overview, annotation pipeline creation |
| Note | Write/delete tools (create_*, delete_*, evaluate_quality, compute_consensus) require AVALA_MCP_ENABLE_MUTATIONS=true. |
Documentation
License
MIT - see LICENSE for details.
