Diagnostics and Log Tools

Diagnostics and log tools read persisted runtime evidence. Omitting fromSeq returns the latest matching events. Supplying fromSeq switches to incremental cursor reads so an agent can continue from the last seen event.

Tool Summary

Tool	Purpose
diagnostic_tail	Read structured LOOMLE diagnostic events.
log_tail	Read Unreal output log events.

diagnostic_tail

Reads structured LOOMLE diagnostic events.

Parameters

Field	Required	Notes
fromSeq	no	Exclusive sequence cursor. Omit for latest matching events; provide to return seq > fromSeq.
limit	no	1 to 1000; defaults to 200.
filters.severity	no	Filter by severity.
filters.category	no	Filter by diagnostic category.
filters.source	no	Filter by source.
filters.assetPathPrefix	no	Filter diagnostics under an asset path prefix.

Use When

Use diagnostics when the issue is likely in LOOMLE’s structured runtime events: tool errors, bridge state, validation failures, or asset-specific diagnostics.

log_tail

Reads Unreal output log events.

Parameters

Field	Required	Notes
fromSeq	no	Exclusive sequence cursor. Omit for latest matching events; provide to return seq > fromSeq.
limit	no	1 to 1000; defaults to 200.
filters.minVerbosity	no	Minimum Unreal log verbosity.
filters.category	no	Single Unreal log category.
filters.categories	no	Multiple Unreal log categories.
filters.source	no	Log source.

Cursor Semantics

Both tools return items in chronological order. When fromSeq is omitted, the page contains the latest limit matching events and nextFromSeq advances to highWatermark for the next polling call. When fromSeq is supplied, the page contains matching events with seq > fromSeq; use nextFromSeq for the next incremental call.

Use When

Use logs when the next clue is likely in Unreal’s output log, such as compile messages, plugin warnings, runtime errors, or editor subsystem output.