// Use with the root `twg` skill for pull-request queues, stale reviews, repo contributors, repo hot areas, PR-based status, issue-to-PR lookups, and engineering review bottleneck analysis.
Use with the root `twg` skill for topic deep dives, dependency maps, graph explanations, workitem/page/user context, project-to-repo discovery, experts, and "catch me up" prompts. Hydrate anchors and relationship evidence before synthesis.
Use with the root `twg` skill for on-call handoffs, HOT/reliability reviews, incident/PIR analysis, Assets and laptop refresh, capacity/staffing views, meeting summaries, and operational risk readouts.
Root operating skill for the TWG CLI. Use for Atlassian work-data tasks across Jira, Confluence, Bitbucket, goals/projects/focus areas, JSM, Assets, Admin, people, teams, org structure, docs, meetings, videos, search, graph context, and supported mutations. Exact command syntax must come from live `twg help`.
Use with the root `twg` skill for personal, team, org, project, goal, focus-area, executive, quarterly, appraisal, and leadership status rollups. Resolve scope first, collect bounded evidence, and synthesize a readout with confidence and gaps.
| name | twg-engineering-work |
| description | Use with the root `twg` skill for pull-request queues, stale reviews, repo contributors, repo hot areas, PR-based status, issue-to-PR lookups, and engineering review bottleneck analysis. |
Use together with the root twg skill. Exact command grammar must come from
live twg help, twg help <terms>, or twg help describe <path>.
Resolve the engineering anchor:
Use typed Bitbucket, Jira, context, and search command families when they clearly match the prompt. Use focused help only when the route or exact contract is unclear.
Query reviewer-scoped open PRs. Sort by waiting time, requested action, unresolved tasks/comments, failing CI, and project relevance. Hydrate only PRs that need action.
Find PRs open or waiting beyond the threshold. Group by repo, author, reviewer, and stage. Identify bottleneck patterns such as missing reviewer, unresolved tasks, failing CI, repeated request-changes, or owner unavailable.
Use workitem context to find linked PRs, commits, branches, and repos. Fetch PRs only if the user asks for details, status, or next action.
Query PRs/commits for the repo and time window. Group by files/areas, authors, reviewers, and themes. For hot areas, prioritize changed area plus frequency and ownership signals.
Resolve org/team first, then collect PRs for members or repos in the time window. Group into themes and repos/services. Call out gaps where PR-only evidence omits Jira, docs, planning, or customer context.