- Upvotes
- 0
- Comments
- 0
- Impressions
- 36
this week
discussions
this week
About
YYLO is an open-source system of work for engineering teams using coding agents. It coordinates work from a defined engineering task through delegated agent execution, validation, review, and accepted code changes. Rather than replacing Git or an existing repository workflow, YYLO makes the process that produces code inspectable. Tasks record bounded outcomes, constraints, dependencies, status, responses, and commit evidence. Agent runs can execute in isolated worktrees while preserving model and session provenance. Validation uses tests, evidence receipts, and policy-bound review so acceptance can be inspected later. YYLO also includes Ledger, which keeps task truth and engineering memory durable in Git, and Benchmark, which compares agent attempts using retained evaluation evidence. The project is designed to preserve the context behind engineering changes so future work can understand not only what code changed, but the intent, attempts, review evidence, and outcomes that led to it.
Problem & solution
Problem
Coding-agent work can lose intent, provenance, validation evidence, and execution history.
Solution
YYLO coordinates agent work and preserves task, session, review, commit, and evaluation evidence.
Who it's for
Engineering teams building software with coding agents.
Main features
- Task truth in Git
- Agent worktree isolation
- Session provenance
- Evidence-based review
- Benchmark attempt comparison
Use cases
- Delegate and validate bounded code changes
- Compare agent attempts with evidence
Articles
1 storyComments
0 shownShare your thoughts about this tool.
Sign in to leave a comment.
No comments yet. Be the first to leave one.