shortlist

shortlist is an open-source library for dynamic tool selection in AI SDK agents. When an agent has 50, 100, or 200+ tools, handing the model all of them on every step is a problem twice over: it burns context tokens, and it lowers tool-calling accuracy — the right tool gets lost in the noise.

shortlist indexes your tools once and, on each step, exposes only the handful that actually match what the user is trying to do.

What Makes It Different

• Zero-config default: Keyword search (BM25 + TF-IDF) that needs no API key and runs in well under a millisecond.
• Better with embeddings: Add an embedding model for semantic search and shortlist fuses it with keyword search automatically.
• Drops into the AI SDK: Use it as a prepareStep, as wrapLanguageModel middleware, or call select() yourself.
• Built for agent loops: Adaptive cutoffs, miss-driven escalation, recently-used tool retention, and related-tool expansion.

It's MIT licensed and free to use — pnpm install shortlist.

github.com/enemyrr/shortlist