Skip to content

ragpicker

ragpicker generates a self-contained, filesystem-based agentskills.io skill that wraps a haiku-rag-built LanceDB RAG database.

Unlike haiku-rag create-skill (which emits a haiku-skills entrypoint package consumed by haiku-rag's own pydantic-ai agent), the skill produced here is driven by an external agent (e.g. Claude Code) that invokes a script. The generated skill provides a PEP 723 wrapper, scripts/haiku_rag.py, exposing search and cite subcommands implemented over the haiku-rag Python client.

How it works

Given a haiku.rag.yaml config and a haiku-rag-built LanceDB database directory (<stem>.lancedb), ragpicker:

  • names the skill from the database directory stem (handbook.lancedb → handbook-haiku-rag);
  • copies the bundled template (shipped as package data under src/ragpicker/template/[dbname]-haiku-rag/), substituting [dbname] with the stem in path names and text-file contents;
  • embeds both the .lancedb database and the haiku.rag.yaml config under the skill's assets/;
  • pins an exact haiku-rag-slim version in the wrapper so the embedded database and the runtime always agree (see Version pinning).

Next steps