Skip to content

RGES-PIT ยท Shared research infrastructure

Microlensing context,on demand.

Nancy is a searchable knowledge base for microlensing software, documentation, tutorials, notebooks, and literature. Connect it to an MCP-capable assistant and ask questions where you already work.

Connect from VS Code Connect from Codex

What you need

  1. A personal Nancy API key. RGES-PIT Slack members can run /mcp_api_key in Slack; other users need an invite code from the Nancy administrators.
  2. An MCP client that supports Streamable HTTP and custom authentication headers.

RGES-PIT Slack members should request their key privately from Nancy:

/mcp_api_key

If you were given an invite code instead, request your personal key once:

curl -X POST https://mcp.rges-pit.com/v2/api-keys/request \
  -H "Content-Type: application/json" \
  -d '{"invite_code":"YOUR_INVITE_CODE","contact":"[email protected]"}'

The response contains api_key. Store it as a password: do not commit it to a repository, paste it into a shared configuration file, or post it in Slack.

Already using the old address?

https://nancy-brain.malpas.nz remains supported. The RGES-PIT address is an additional stable name for the same service.

What Nancy can do

  • Search the microlensing knowledge base semantically.
  • Retrieve precise passages with source and line metadata.
  • Explore indexed document and repository trees.
  • Apply personal retrieval weights to prioritize useful sources.
  • Report service and index status.

Nancy exposes knowledge-base tools, not a language model. Your chosen AI client decides how to use the returned evidence and generates the answer.

Building your own knowledge base

This site documents the shared microlensing service. To build, configure, or self-host a different knowledge base, use the Nancy Brain package documentation.