Immutable plans
Bind cases, models, attempts, source identities, and constraints before execution.
Evaluation evidence / longitudinal comparison · v0.1.0-rc.7
Compare agent attempts with evidence that lasts.
YYLO Benchmark validates immutable plans, runs isolated attempts, governs grading, reconciles evidence, and produces bounded reports without hiding source artifacts.
Supplementary visual overview
YYLO Benchmark at a glance.
Product behavior and commands are defined by the verified HTML documentation; this artwork provides a high-level visual summary.
Open full-size image
Install / first run
Start with one bounded outcome.
Install the synchronized release, then begin with a change small enough to validate and review.
Available today / mechanisms
Control is visible, not implied.
Isolated attempts
Keep candidate runs independent and retain execution and cost provenance when available.
Governed grading
Use pinned graders and blinded evaluation rather than untraceable model preference.
Recovery and reports
Reconcile durable intent, recover safely, and compare retained evidence over time.
Product direction / clearly separated
How this layer strengthens the System of Work.
Benchmark evidence is designed to help teams learn which agent and workflow works best for a repository and change type. Automatic production workflow selection requires connected real-world outcome history and is not a current-release claim.