Try Benzi- A coding agwnt that _queries_ your codebase instead of reading it
Summary
Benzi is a coding agent that queries codebases to understand and write code, achieving 77.4% on SWE-bench Verified for under $30, and is model agnostic.
Similar Articles
@jasonzhou1993: I gave my coding agent a map across 3 of my repos: - Reduce ~50% token - a grep hook so every grep call has much richer…
Jason Zhou shares a technique to give coding agents a map across repos, reducing token usage by ~50% and enabling richer grep and call chain tracing. He also releases a Claude Code plugin marketplace called AI Builder Club Skills for setting up codebase harness and compounding agent loops.
the agent that stuck for me writes zero code, it just builds my friday sprint review
An AI agent that writes no code but helps build Friday sprint reviews, highlighting a specific productivity use case.
SWE-Explore: Benchmarking How Coding Agents Explore Repositories
SWE-Explore introduces a benchmark for evaluating coding agents' repository exploration capabilities, requiring ranked lists of relevant code regions within line budgets. Experiments show agentic exploration outperforms traditional retrieval, and line-level coverage remains a key differentiator.
@svpino: Claude Code to write your code and Codex to verify it. I met with a team that's been doing this for a few weeks now. I …
A developer shares an approach where teams use Claude Code for writing code and Codex for verification, focusing on detailed specs and overnight AI agent runs.
@_akhaliq: SWE-Bench ProMax Benchmarking Agents on Large-Scale Multilingual Code Refactoring paper: https://huggingface.co/papers/…
Introduces SWE-Bench ProMax, a multilingual code refactoring benchmark of 170 instances across 7 programming languages to evaluate AI coding agents. Frontier models achieve only 41.2% resolve rate, confirming it as a challenging and unsaturated benchmark.