Verified Platform Briefing — 100-Step Publish, Prove, Recover

Failed
59.6s119 persisted spans10 services0 tools9 errorsmodel MichaelAI (default origin)Jul 20, 03:05:49 PM UTC

Entire trace context

Intent

A real 100-step distributed agent publishing workflow with inspectable operational decisions, cross-site propagation, durable mutations, readbacks, ASCII transformation, and deASCII reconstruction.

Input

No public-safe input or prompt name was recorded.

Output

{"verdict":"failed","artifactCount":4,"requiredFailed":8}

Evidence

{"outcome":"created","wordCount":648,"contentHash":"14c00d81b5120e324a0dec7790f3a2453efbd426b0b9862e43f3a32d38e9aa10","entityIdHash":"b0ab999fb10d2a96","changedFields":[],"versionCreated":false,"redactionStatus":"public_safe_summary_only","privateNoteContentIncluded":false} · 14 linked items (markdown, artifact, http_response, json, trace_snapshot, eval_result) · 338 events

Model telemetryMichaelAI (default origin)
Semantic names
59.62 / 59.62
Execution
00.0014.9029.8144.7159.62
Selected operation
Select an operation

Select an operation in the story, tree, waterfall, or graph.

Final output
{
  "verdict": "failed",
  "artifactCount": 4,
  "requiredFailed": 8
}