Repo · Search & RAG

Azdaja

Recursive language-model layer holding full inputs in a local evaluator and exposing a small working surface, with optional typed semantic judgments…

Open github.com ↗

From the repository

Minimal harness-agnostic recursive language model layer — one binary, Python + llm()

agent-skillsclaude-codecodex-clicontext-engineeringcontext-windowdeveloper-toolsgemini-cligemini-cli-extensionjcodelarge-language-modelsllm-toolslong-contextopencoderecursive-language-modelsrust

How builders describe it

Recursive language-model layer holding full inputs in a local evaluator and exposing a small working surface, with optional typed semantic judgments alongside generative calls.

Azdaja is a bare OSS RLM layer you can plug into Claude, Codex, OpenCode and others now with Jev, branches that only need a judgment can return one directly, useful for reranking, verification, classification and semantic joins, cutting unnecessary model work

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