Stage Payload Reverts

Payload regression reverts are automated end-to-end by orchestrating Jira bug creation, revert pull requests, and payload validation job triggers across multiple tools. This Skill coordinates the full workflow using GitHub CLI and Jira access.

Before triggering an automated payload revert workflow, you need GitHub CLI, Jira MCP access, and repository write permissions. You must also provide context in memory including results_yaml_path, payload_tag, version, stream, architecture, release_controller_url, and revert_candidates.

Yes, you can process multiple revert candidates in a single regression triage run. The Skill handles parallel candidate processing, providing per-candidate outcomes and error handling to manage multiple identified regressions simultaneously.

Jira bug creation for unresolved payload regressions works by automatically creating TRT Jira bugs in an idempotent manner. This ensures duplicate bugs are not created when coordinating revert staging across CI tools and validation jobs.

The revert process synchronizes state across CI tools by updating payload results YAML and HTML reports with revert status and links. It uses the payload-results-yaml and payload-autodl-json schemas to coordinate state throughout the orchestration.

You should not use automated payload revert orchestration when revert candidates lack high confidence identification from analyze-payload. The workflow requires high confidence candidates, proper schemas, and necessary tool access to coordinate state safely across CI systems.