loopbench rtre84 · COMPLETED
Benchmark agentic loop quality of real harness/model combos: how well does a model+harness iterate — run tools, observe, diagnose, self-correct — until a deterministically verifiable goal is met?
github.com/rtre84/loopbench · ★ 0 · Forks 0 · Size 32 KB
SUMMARY
Technologies 5
Scored 5
Observed 5
Practices 6
Evidence 5
Skips 0
COVERAGE
Analyzed 66 files · 2 commits · 0 API calls
TECHNOLOGIES & DEPTH
Shell LANGUAGE Depth 70
10 files · PRODUCTION
JSON LANGUAGE Depth 70
2 files · PRODUCTION
Markdown LANGUAGE Depth 70
3 files · PRODUCTION
TOML LANGUAGE Depth 70
7 files · PRODUCTION
Python LANGUAGE Depth 70
39 files · PRODUCTION
PRACTICES
documentation · observedautomated_tests · observedcontinuous_integration · absentcontainerization · absentlinting · absentformatting · absent
ACTIVITY & OWNERSHIP
First commit 2026-08-04
Last commit 2026-09-08
Active months 2
Commits 2