uat-executorlisted

# UAT Executor An L2 Truth-Execution skill. Where `e2e-test-writer` **produces** test files, this skill **runs** scenarios against a real environment and records what happened. The output is `uat-raw-report.md` — the raw material that `test-result-analyzer`, `observability-analyzer`, and `release-decision-engine` all consume to make the GO/CONDITIONAL/BLOCKED call. Lying in that raw material poisons every downstream decision, so the skill's whole job is "run honestly and capture evidence". ## When You're Invoked - During PIPELINE-3 step 3, after `e2e-test-writer` has landed automated specs and the environment is ready for UAT. - On demand as `/vibeflow:uat-executor <env> <scenario-glob>`. - From `release-decision-engine` before a CONDITIONAL release, when a fresh UAT run is required before sign-off. ## Input Contract | Input | Required | Notes | |-------|----------|-------| | `scenario-set.md` | yes | Output of `test-strategy-planner`. Only scenarios tagged `uat` or `e2e+uat` are executed; others are owned by different skills. | | Target environment | yes | `--env <name>` — must resolve to a URL in `vibeflow.config.json.environments.<name>`. `production` is rejected at the precondition stage (UAT against production is not UAT). | | `test-strategy.md` | optional | When present, informs per-scenario retry policy and halt policy overrides. | | Runner availability | derived | `playwright` / `detox` via `repo-fingerprint.json`. Automated steps require the matching runner;