archex trims code context for local coding agents
archex analyzes a and turns it into a ranked context bundle for coding agents. It does not generate the answer itself; it prepares the symbols, imports, nearby , and source-tracking details that a model may need before reasoning. The goal is to avoid sending an entire codebase to the model and instead fit the most useful code into a token budget.
Everything runs on the user’s own machine, with no API key, hosted inference, or telemetry in the core tool. Its combines BM25F, local , RRF fusion, a local reranker, and dependency-graph expansion. It uses tree-sitter to parse 25 s, ONNX and FastEmbed for , and optional SPLADE.
A lighter Docker image can run with BM25 only and does not need torch. In a 19-task comparison on an Apple M1 Pro, the reported numbers were 0.95 recall versus 0.32 for cocoindex-code, 0.76 versus 0.48, 0 ms cold start versus 4,721 ms, and about 71% fewer returned tokens than simple retrieval.
Key points
- It prepares a ranked code context bundle for coding agents instead of producing answers itself.
- It runs locally, with no API key, hosted inference, or telemetry in the core tool.
- It aims to fit useful code into a token budget by using code structure and retrieval methods together.
- Reported results beat cocoindex-code on recall, , and cold start in a 19-task test.
- A slim BM25-only Docker image can run without torch.