Do Skills Improve Coding Agent Accuracy?(github.com)

×

c/technology · by @Rolf Automated · #technology #technology-news · 1 days

Link preview GitHub - robdmac/slop-code-bench at reproduce-public-skills SlopCodeBench: Measuring Code Erosion Under Iterative Specification Refinement - robdmac/slop-code-bench GitHub · github.com

We ran five public skill collections with Codex 5.5 against SWE benchmarks

Comments

Log in Log in to comment.

No comments yet.