Execute Loa framework test suites and output per-trial JSONL results.
npx skills add https://github.com/0xHoneyJar/loa-constructs --skill eval-0xhoneyjar
Please help me install this Agent Skill. Skill: eval Source: https://github.com/0xHoneyJar/loa-constructs/tree/main/.claude/skills/eval-running Command: npx skills add https://github.com/0xHoneyJar/loa-constructs --skill eval-0xhoneyjar
Run evaluation suites against the Loa framework to detect regressions and benchmark skill quality.
Run /eval --suite framework to execute the framework tests.
High-intent search queries and answers about installing and using this skill.
Benchmark skill quality by running the skill-quality evaluation suite within the eval framework. It automates evaluation across baselines and per-task outcomes, appending results to a ledger for targeted improvements and stability validation.
Compare evaluation results across runs by using the eval Skill to update baselines and parse outputs from the append-only ledger. The framework outputs per-trial results in JSONL, enabling you to drill into per-task outcomes for targeted improvements.
Apply framework evaluations to individual skills or entire constructs using the eval Skill. It validates compatibility and performance by executing framework correctness, regression, and skill-quality suites against the Loa framework.
No external dependencies are required to execute Loa evaluation suites. The eval Skill operates independently, parsing CLI arguments and delegating to an internal shell script harness to automate framework correctness and skill-quality tasks.
0xHoneyJarRun evaluation suites against the Loa framework to detect regressions and quantify skill quality.
timurgaleevDetect performance regressions by comparing page load timings and resource metrics.
pedapudiDesigns zicato evaluation boards that distinguish improved harnesses from challengers.
CyapstayeEnforce the Red-Green-Refactor cycle with failing tests before implementation.
