AI Skills & Agentstjcageslinear-sync
linear-sync
Audit and rescue an existing Linear project by reconciling its issues, milestones, dependencies, and tracking protocol with repository reality.
Installation
npx @compound-design/skills get tjcages/linear-syncContent
Linear sync
Chat rules: RESPONSE.md. Auth: AUTH.md. Playbooks: METHODOLOGY.md §1a, §18, §21–§22, §26. Examples: EXAMPLES.md.
Step 0 — auth
Probe Linear MCP. On failure → AUTH.md; stop.
Operating order
- Identify project — search by name; confirm with user if multiple.
- Enumerate — open+closed issues, milestones, labels, documents, status updates (§1a).
- Score — §18 rubric (0–2 × 10). Pass ≥16/20 with no load-bearing 0. Emit gap list.
- Confirm — per-category yes/no before writes. Reorganize is higher-risk than extend.
- Fix in order (§22a): orphans → milestones → milestones spam collapse → dependencies → install/update discipline protocol (§5).
- Half-tracked? — scan
docs/for unmirrored roadmaps (§22b); separate milestone sets. - Backfill-late? — §22c; corroborate doc claims (§4.8).
- One status update recording the cleanup if health/structure changed (§20).
- Re-score — show before/after totals. One next action.
Hard rules
- Search before every create.
- Never delete/reorganize off-limits or other-owned issues without explicit consent.
- Ticket prose stays terse (§4). Chat follows RESPONSE.md.
Related
Written by tjcages in tjcages/linear, under the MIT licence. Source: https://github.com/tjcages/linear/blob/main/skills/linear-sync/SKILL.md