What problem does it solve? Maintainers need a controlled, auditable way to run end-to-end test suites against release candidates without accidentally triggering the wrong workflow mode, testing the wrong commit, or accepting incomplete evidence as a passing result. ## Core Features & Use Cases - Mode Selection: Maps maintainer requests to ordinary, Launchable, or full E2E modes with strict rules that prevent a generic request from authorizing the Brev Launchable path. - Trusted Dispatch: Resolves the candidate SHA from origin/main, generates a unique correlation ID per run, and dispatches the trusted e2e.yaml workflow via the GitHub CLI. - Evidence Verification: Collects run and job JSON, validates the Launchable E2E artifacts with a validator script, and binds successful full runs as release evidence for the release-tag workflow. - Use Case: Before cutting a release tag, a maintainer asks to run the full E2E suite; the skill dispatches the workflow against the exact candidate SHA, waits for completion, validates the Launchable evidence and cleanup receipt, and returns the verified run URLs for the release ledger. ## Quick Start Ask the agent to run the full E2E suite for the current release candidate on origin/main and return the verified workflow run and Launchable job URLs.
