agent-satisfaction

name	agent-satisfaction
description	Score all agents across the 17 satisfaction intents. Generates a comprehensive report with per-agent scores (1-10) and system-level scores.
version	1.0.0
author	ops-officer
tags	["agent","satisfaction","health","wellbeing","team","scores"]
trigger	{"command":"/agent-satisfaction","keywords":["agent satisfaction","team health","agent scores","wellbeing check"]}

agent-satisfaction

Score all agents across the 17 satisfaction intents.

Procedure

1. Run the scoring script:

   bash /home/node/.openclaw/scripts/agent-satisfaction-score.sh
   

2. Review the output — scores each agent on 11 per-agent intents and 6 system intents

3. Identify any scores below 6 — these need attention

4. For low scores, propose specific actions:

   • Focus < 6: Propose agent split
   • Voice < 5: Trace pipeline, find drop point
   • Harmony < 5: Identify friction source
   • Equipped < 6: Seed missing charts, create skills
   • Clarity < 5: Rewrite SOUL.md
   • Trusted < 5: Investigate completion rate, handoff reliability

5. Store the report in Chartroom:

   memory_store report-agent-satisfaction "<summary>" --importance 1.0 --category fact
   

6. Post summary to #ops-dashboard if any agent scores critically low (avg < 5)

Output

Full ASCII report with bar charts, per-intent scores, and action recommendations.

Intent: Observable [I13]. Purpose: [P-TBD].