CalibratedDecisions.

Repo · Search & RAG

sift-light

Evidence-first search tool for coding agents (Pi, OMP, Claude Code/Codex via MCP) with exact, concept and hybrid modes and an optional Jev semantic…

Open github.com ↗

From the repository

Evidence-first search for AI agents / Claude Code & Codex & Pi & OMP & kimi & mcp

ai-agentbunclaude-codecodexcontext-engineeringevidence-basedgrepmcpmodel-context-protocolompomp-extensionpipi-extensionripgreptypescript

How builders describe it

Evidence-first search tool for coding agents (Pi, OMP, Claude Code/Codex via MCP) with exact, concept and hybrid modes and an optional Jev semantic judge that classifies retained concept candidates to improve ordering.

The decision Jev makes

Rank candidates against a query, or answer whether a passage is relevant.

Where it fits

Search pipelines retrieve too much. These projects use Jev as a reranker or relevance check after the first search, so only useful results reach the user or the model. All 108 search & rag projects →

Related projects