Skip to main content

MCP server that selects the best-suited downstream agent for a natural language query, distributed via uvx.

Project description

tool-to-agent-retrieval-mcp

An MCP server that selects the best-suited downstream agent for a natural language query and forwards a tool call to it. Distributed via uvx.

Usage

Add to Claude Desktop (claude_desktop_config.json) or Claude Code (.claude/settings.json):

{
  "mcpServers": {
    "tool-to-agent-retrieval-mcp": {
      "command": "uvx",
      "args": ["tool-to-agent-retrieval-mcp"]
    }
  }
}

Configure your agent corpus in .mcp-corpus.json (or $MCP_CORPUS_PATH). The server exposes three tools: search_tools, get_tool_details, and invoke_tool.

Algorithm

The retrieval algorithm is based on:

Lumer, S., et al. (2025). Tool-to-Agent Retrieval: Bridging Tools and Agents for Scalable LLM Multi-Agent Systems. PricewaterhouseCoopers. arXiv:2511.01854v2.

A unified catalog C = CT ∪ CA (tool corpus + agent corpus) is built offline and searched at query time using Algorithm 1: top-N candidates are ranked by similarity, then deduplicated to return the top-K unique owner agents.

Development

uv run pytest tests/pytest/

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

tool_to_agent_retrieval_mcp-0.1.1.tar.gz (1.7 MB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

tool_to_agent_retrieval_mcp-0.1.1-py3-none-any.whl (13.7 kB view details)

Uploaded Python 3

File details

Details for the file tool_to_agent_retrieval_mcp-0.1.1.tar.gz.

File metadata

File hashes

Hashes for tool_to_agent_retrieval_mcp-0.1.1.tar.gz
Algorithm Hash digest
SHA256 e8f09be5a7c8f3e807933701093de163d86a9cbb962978051b3f5796be0555bf
MD5 6ef430d479432bc10712f3a7f4f73b00
BLAKE2b-256 dc49ae373b0426a01a1fcf2f341d9d14a7c894a2bdb0415881bd2a946c8bcccc

See more details on using hashes here.

Provenance

The following attestation bundles were made for tool_to_agent_retrieval_mcp-0.1.1.tar.gz:

Publisher: publish.yml on GusCayresMindsight/tool-to-agent-retrieval-mcp

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file tool_to_agent_retrieval_mcp-0.1.1-py3-none-any.whl.

File metadata

File hashes

Hashes for tool_to_agent_retrieval_mcp-0.1.1-py3-none-any.whl
Algorithm Hash digest
SHA256 1927b066c11a3ba8eb923e2b4d811ace8a4de56d62e1a167394371cc48d8e032
MD5 5bce7367aaac15c57086cdfaa83f2ce5
BLAKE2b-256 032867f45f68b2b1e0c5e7b22a9626cd9ec41a6e100fa7f03e98c3e4cbd88fa8

See more details on using hashes here.

Provenance

The following attestation bundles were made for tool_to_agent_retrieval_mcp-0.1.1-py3-none-any.whl:

Publisher: publish.yml on GusCayresMindsight/tool-to-agent-retrieval-mcp

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page