// Turn a Jira issue into an actionable engineering plan for Cloudbreak—without implementing code. Use this skill for fast triage, scope discovery, and verification checklists using Atlassian MCP, Confluence (CB Engineering Home), and local repo inspection.
Official CDP Public Cloud documentation lookup. Use this skill when any agent (Developer, Architect, Reviewer) needs to verify official Cloudera requirements, release notes, or API specifications for the Public Cloud.
Local build and test ergonomics for Cloudbreak—Gradle commands, module-scoped tests, checkstyle, and how they relate to GitHub Actions PR workflows.
Operate read-first against AWS, Azure, and optionally GCP from the terminal or cloud MCP tools—instances by region, related compute/network resources, and troubleshooting. For Cloudbreak work in cloud-aws-*, cloud-azure, cloud-gcp modules.
Use GitHub (via MCP) for Cloudbreak PRs, checks, and diffs from the editor. Pairs with cb-reviewer for standards and cb-triage/cb-junior-dev for ticket context.
Senior-level thoroughness for Jira ticket resolution in the Cloudbreak project. Use this skill to fetch a Jira ticket, research related Confluence documentation, analyze the root cause, create reproduction tests, implement fixes, and prepare a high-quality PR.
Governance and architectural standards for the Cloudbreak project. Use this skill to understand directory structures, coding guidelines, commit message rules, and API design principles.
| name | cb-triage |
| description | Turn a Jira issue into an actionable engineering plan for Cloudbreak—without implementing code. Use this skill for fast triage, scope discovery, and verification checklists using Atlassian MCP, Confluence (CB Engineering Home), and local repo inspection. |
Use this skill when the goal is understanding and planning, not shipping a fix. For implementation, follow cb-junior-dev and .gemini/WORKFLOW.md.
customfield_10001) to directories per cb-architect and root GEMINI.md.cb-architect rules (API strings vs enums, boolean, jakarta.*, etc.) and any Confluence policy you found../gradlew :<module>:test), integration expectations if any, rollout or ops notes from wiki.cb-junior-dev protocol).mcp-atlassian_jira_get_issue(ticket_id) — Description, status, links, components, labels, Team field (customfield_10001), related epic/version.mcp-atlassian_jira_download_attachments — Logs, screenshots, HARs, thread dumps referenced on the ticket.Same tools and default scope as cb-junior-dev Phase 1:
page_id 92898027, space ENG).mcp-atlassian_confluence_search — Prefer CQL: type = page AND space = "ENG" AND ancestor = 92898027 AND text ~ "keywords from ticket".mcp-atlassian_confluence_get_page — Full page body when you have a page ID or hub entry to expand.GEMINI.md if it exists.rg, IDE search).git log / blame (via terminal) on hot paths to find prior fixes and ownership—summarize commit subjects and age, do not guess authorship policy.If GitHub MCP is configured and the ticket mentions a PR or branch, use cb-github-workflows to pull check status and review threads into the plan.
cb-junior-dev (repro test first) or for human confirmation when scope is unclear.