Local Indexing to reduce tool calling and token usage

Pensieve is a local indexing tool that reduces tool calling and token usage for coding agents by providing semantic and exact-match search across files, repositories, and conversations. Connected via MCP (Model Context Protocol), it delivers precise context in fewer calls—averaging 2.0 tool calls per lookup compared to 3.4 with grep—helping agents find relevant code passages with file and line numbers on the first attempt. The tool integrates with AI coding assistants like Claude, Codex, and Cursor to enable efficient "search before you grep" workflows.