Skip to content
AI usage dashboard

synthetic-legacy-archive expectations (synthetic)

Legacy JSONL under agents/main/sessions/ uses placeholder normalized usage fields from documentation: input_tokens/output_tokens, prompt_tokens/completion_tokens aliases and cacheRead. sessions/sessions.json is another legacy migration input.

Manually calculated expectations:

  • Discover both files with instance root=agents/main.
  • Treat JSONL as migration/offline maintenance input. Official documentation says Gateway startup does not import it and openclaw doctor –fix must migrate it; entry schema is undocumented. Return unknown_format with a pending-verification reason. sessions.json likewise returns unknown_format, identifying the doctor migration requirement.
  • No usage_events/source_aggregates. Synthetic values such as 1200+300/60+15 do not count; native schema is unverified, with no guessed fields or zero usage.
  • Repeat scans add no data.