Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.selftune.dev/llms.txt

Use this file to discover all available pages before exploring further.

Usage

selftune last
Shows a diagnostic summary of the most recent agent session, including:
  • Which skills fired
  • Which queries went unmatched
  • Any improvement signals detected
  • Suggested actions

Grading output

When grading data is available, selftune last also displays:
FieldDescription
Overall gradepass, partial, or fail
Tier 1 (trigger)Score 0.01.0 — did the right skill fire?
Tier 2 (process)Score 0.01.0 — was the execution correct?
Tier 3 (output)Score 0.01.0 — was the output useful?
See Grading concepts for how tiers map to the overall grade.