📖About & Usage

About

Obi-Wan UAT is an automated acceptance tester that independently verifies Leo-owned blueprint initiatives once they pass engineering review. It exists so that initiatives can close on real acceptance rather than waiting on a human sign-off, keeping the build pipeline moving without bottlenecks. The droid runs every 15 minutes, checking initiatives whose acceptor routes to Obi-Wan (the owner agent) against their success criteria.

When ob-uat-01 activates, it selects reviewing initiatives assigned to Obi-Wan as acceptor, reads the initiative state through the API/relay (never touching a decoy database), and runs an LLM-powered acceptance check. It compares the delivered results with the stated success criteria, looking for clear evidence that each must_have requirement is met. A pass triggers a pipeline_uat action to close the initiative; a fail sends it back for rework, capping re-tries at three.

This droid is a unique, one-off member of the uat family—no other agents run copies of this family. Its execution engine isn’t recorded in Carolverse sources, so we don’t know for sure whether it’s pure software or calls Claude. Given its duty to run LLM acceptance checks, it’s likely that a language model is invoked under the hood, but the specifics (model, integration) remain undocumented.

🛰️Updates

Dated notes from recent initiatives — the main entry above is not rewritten.

👤Owner

📚Recent initiatives

Initiatives that touched this droid — a short summary each; open one for the full story.

No initiatives recorded for this droid yet.