Skip to content

armgit / xerj

Public Imported from GitHub
0

AI search engine and MCP server for bots and agents. One Rust binary indexes code, docs, logs, JSONL/CSV and PDFs for BM25 full-text, vector and hybrid search, code search without grep, agent memory and RAG. Elasticsearch-compatible API; Corpus Hub serves hundreds of reference memory corpora.

697 B
Raw Blame

description: Search your indexed local documents and answer with the file and page each fact came from argument-hint:

Answer this from the user's indexed local documents: $ARGUMENTS

  1. Search with the xerj_search MCP tool (or xerj search --prefix <prefix> "<plain words>"), using the words the documents would use. If you know the prefix the folder was indexed under, pass it. Try two or three phrasings if the first one misses.
  2. Answer only from what the hits say. Cite each fact with the file path and the page, slide, sheet or message the hit names.
  3. If nothing relevant is found, say so and say which prefix was searched. Do not answer from memory.