session-status-briefinglisted

## When to use Invoke at the start of a work session on a familiar repo when you need a full orientation snapshot: what changed, what's risky, what's dead, and whether the retrieval stack is healthy. Trigger phrases: "what's the status", "bring me up to speed", "what changed since last time", "status check", or any session-open where the user hasn't specified a task yet. ## Steps Run steps 1, 2, and 3 in parallel — they are independent. ### 1. Read HANDOFF.md Read `HANDOFF.md` from the project root. Extract: - Current state (what's broken, what's in progress) - Single most important thing to know - Any known blockers or environment quirks If HANDOFF says something is fixed but symptoms persist, trust the symptoms over the doc. ### 2. Stack health check Run the health check script to get ground truth on stack state: ```bash bash healthcheck.sh --quick 2>&1 | grep -E '^\s+(OK|X|WARN|CRIT)|^HEALTHCHECK' ``` (The `grep` keeps the full OK/X line list — `tail` alone truncates the failures off the bottom.) Look for `HEALTHCHECK: ok` vs `HEALTHCHECK: fail`. Flag any CRIT items prominently before the rest of the briefing. Common failures: - `mcp-servers-down(duckdb)`: duckdb cold-start — self-heals on first query; not actionable. - `stack-not-at-head`: packages behind HEAD — fix: `stack-not-at-head-remediation` skill. - `jcodemunch-stale`: index stale — fix: `bash scripts/jcodemunch-reindex.sh` (self-heals the local/git dual-identity collision and retries). The "X not Con