YYYYLO
A command-line orchestrator for coding agents that gives developers a quick agent loop and operators typed task, validation, merge, and release-readiness boundaries with receipt-backed evidence.
Open-source projects by this author in the catalog, across all categories and statuses.
2 Projects
YYA command-line orchestrator for coding agents that gives developers a quick agent loop and operators typed task, validation, merge, and release-readiness boundaries with receipt-backed evidence.
YYYYLO Benchmark is the evaluation companion to the YYLO orchestrator — a trusted-host experiment runner for historical Ledger tasks, supplied coding prompts, and workflows. Each candidate attempt runs in a fresh repository with default exclusions for answer-bearing files, outputs are retained for later re-evaluation with new checks or judges, and frozen project/task checklists plus deterministic loss give reproducible comparison rows. It deliberately does not choose a winner or combine judge opinions with test results.