Skip to content
IndexDesign corpus

Agent run history

AI & Agents · 7 checks · 13 resources

The retrospective log that makes an autonomous run auditable after the fact.

Before you ship7 checks
  1. 01Run list with trigger, start time, duration, outcome and cost
  2. 02Per-run step trace — each tool call with its inputs, outputs and errors, expandable without leaving the page
  3. 03Artefacts produced linked from the run — files written, records changed, messages sent
  4. 04Failure classification — distinguishing a model error, a tool error, a permission denial and a budget exhaustion
  5. 05Replay or resume — re-run from the start or continue from the failed step with corrected input
  6. 06Retention and export — how long traces are kept, and a way to export one for a support conversation
  7. 07Filtering by agent, status and date range
Resources that help you build this13