linear-agent-skillslisted

# Linear Agent Skills & Automation Suite (August 2026 Edition) This skill equips coding agents (Antigravity, Claude Code, Cursor, Codex, Gemini, Herdr) to interact autonomously with the Linear Agent interface (`https://linear.app/docs/linear-agent`), manage issue locks, record resolution cycle metrics, generate automated project health reports, and auto-promote retrospective learnings to ThumbGate RAG memory. --- ## Skill Components ### 1. `linear-agent-closeout` **Trigger**: When closing or marking a Linear issue as `Done`. **Command**: ```bash node tools/linear-agent-bridge.js --done <ISSUE_ID> \ --agent <PRIMARY_AGENT> \ --co-agents <SECONDARY_AGENTS> \ --duration "<CYCLE_TIME>" \ --bottleneck "<BOTTLENECK_DESCRIPTION>" \ --improvement "<ACTIONABLE_STRATEGY>" \ --comment "<PR_OR_SHA_PROOF>" ``` **Effect**: - Sets Linear issue state to `Done`. - Attaches labels: `agent:<primary>`, `agent:<co-agent>`, and `agents-multi` (if multiple agents). - Posts structured **Agent Completion & Cycle Time Report** markdown comment. - Auto-promotes `--improvement` strategy to **ThumbGate RAG memory** via `npx thumbgate capture`. - Writes mirror claim note in Obsidian Vault (`~/Documents/AI-Agent-Sync/Handoffs/linear-claims/`). --- ### 2. `linear-agent-project-health` **Trigger**: At session end, weekly review, or when updating Linear project status. **Command**: ```bash node tools/linear-project-update.js ``` **Effect**: - Queries live GitHub open PR count, auto-merge que