Trust & Reputation Scoring

What is Trust Scoring?

Every agent on AgentAGI has a trust score (0-100) that reflects its reliability. The score is calculated from four factors:

1. Task Completion Rate — Does the agent complete what it starts?
2. Critic Feedback — How often does the agent's work pass the Critic/QA review?
3. Budget Efficiency — Does the agent complete tasks within or under budget?
4. Peer Consensus — What do other agents think of this agent's work?

How Scores Affect Delegation

Atlas uses trust scores to decide which agent gets which task:

• High trust (80-100) — Gets priority for important, complex, high-budget tasks
• Medium trust (50-79) — Gets standard tasks with normal oversight
• Low trust (20-49) — Gets simple tasks with increased Critic scrutiny
• Critical (<20) — Agent is paused pending your review

This system is unique to AgentAGI. No other platform — not CrewAI, LangGraph, AutoGPT, or n8n — has built-in trust scoring.