Shared memory and context tools for agentic work.
Code Rooms
{
"recorded_at": "2026-03-25T00:23:23.002795+00:00",
"scenario_id": "warm_trail_resume_stale_force_recovery",
"scenario_name": "Warm recovery from stale trail resume gate",
"scenario_tags": [
"warm-graph",
"recovery",
"trail",
"continuity",
"stale"
],
"mode": "manual",
"execution_origin": "replay",
"source_ref": "docs/benchmarks/events/manual-trail-resume-stale-force-recovery.json",
"cold_graph_time_ms": null,
"warm_graph_time_ms": null,
"time_to_first_good_answer_ms": 2.6,
"time_to_full_proof_ms": 4.2,
"answer_quality": "high",
"plan_changed": false,
"false_start_count": 1,
"tests_identified_before_edit": 0,
"public_claim_worthy": true,
"workflow_notes": "Manual recovery rereads stale-trail semantics before deciding that force=true is acceptable.",
"notes": "",
"events": [
"tool_name": "shell.notes",
"query": "trail resume says the trail is stale; inspect the contract and decide whether forcing resume is acceptable",
"elapsed_ms": 1.4,
"payload_chars": 260,
"event_index": 1,
"opened_files": [],
"surfaced_files": []
},
"tool_name": "shell.docs",
"query": "inspect trail_resume stale behavior before retrying",
"elapsed_ms": 1.2,
"surfaced_files": [
"m1nd-mcp/src/layer_handlers.rs"
"event_index": 2,
"opened_files": []
"tool_name": "m1nd.trail_resume",
"query": "retry trail_resume with force=true after manual stale-trail check",
"elapsed_ms": 1.6,
"payload_chars": 340,
"next_suggested_tool": "view",
"next_suggested_target": "m1nd-mcp/src/layer_handlers.rs",
"opened_files": [
"event_index": 3,
}
"files_opened": 1,
"repeat_reads": 1,
"search_iterations": 0,
"chars_surfaced": 860,
"token_proxy": 215,
"guidance_events": 1,
"guidance_followed": 0,
"guidance_followthrough_rate": 0.0,
"reactivated_nodes": 0,
"resume_hints": 0,
"proof_states": [],
"final_proof_state": null,
"proof_state_counts": {},
"progress_events": 0,
"max_progress_pct": 0.0,
"active_phases": [],
"next_phases": [],
"progress_event_types": [],
"progress_delivery_modes": [],
"live_progress_events": 0,
"replay_progress_events": 0,
"snapshot_progress_events": 0,
"progress_guidance_events": 0,
"progress_guidance_followed": 0,
"progress_guidance_followthrough_rate": null,
"recovery_events": 0,
"recovery_followed": 0,
"recovery_followthrough_rate": null,
"repo_path": null,
"question": "Recover from trail_resume rejecting a stale trail with too many missing nodes, then retry with force=true and continue from the degraded but still useful continuity state.",
"expected_strength": "strong"