npx claudepluginhub otmo123/opentesters/replaySearches pro-workflow database for task-relevant past learnings and session history using FTS5 BM25, producing ranked briefing with corrections, stats, and suggestions.
/replayFull session replay showing thinking blocks, tools, and results in sequence
Re-run agents from a previous run, optionally filtering to only failed agents or specific roles.
/opentesters:replay <runId> [--failed-only] [--role=X]
Examples:
/opentesters:replay run_1709478123/opentesters:replay run_1709478123 --failed-only/opentesters:replay run_1709478123 --role=payment-failure-user| Flag | Default | Description |
|---|---|---|
--failed-only | false | Only re-run agents that failed in the original run |
--role=X | all | Only re-run agents with this role |
mcp__opentesters__get_run_status with runId to get the original run state--failed-only: filter to agents with status failed--role=X: filter to agents matching the rolemcp__opentesters__run_agents with the filtered decompositionmcp__opentesters__synthesize_issueReplay: run_1709478123 → run_1709478456
Original: 10 agents, 7 passed, 3 failed
Replay: 3 agents (failed-only), 2 passed, 1 failed
Newly passing:
✅ AC-002: Supabase unique constraint — now passes
✅ AC-005: Sentry alert on duplicate — now passes
Still failing:
❌ AC-003: Button debounce — still fails (timing issue)