Status
Live
Production-ready
Strategic Intelligence
Risk Evaluator Agent
Provides deterministic risk and confidence scoring for autonomous action gating. Computes a composite risk score (blast radius, reversibility, data confidence) and a confidence score (data quality, pattern match, histori
Intelligence & AnalyticsLiveInternal (Colaberry Enterprise)Verified
Department
Strategic Intelligence
Intelligence & Analytics department for Colaberry Enterprise agents
Source
Internal (Colaberry Enterprise)
Built by Colaberry
About
About the Agent
What this agent does, the challenges it addresses, and where it delivers value.
Provides deterministic risk and confidence scoring for autonomous action gating. Computes a composite risk score (blast radius, reversibility, data confidence) and a confidence score (data quality, pattern match, historical success) to decide whether an action can auto-execute or must be proposed for human review.
Challenges This Agent Addresses
- 1**Operations**: Ensures low-risk, high-confidence agent fixes execute automatically while uncertain changes require human approval
- 2**Risk Management**: Provides transparent scoring breakdown for every autonomous decision
- 3**Compliance**: Governance thresholds are configurable via database, enabling policy adjustments without code changes
Workflow
How the Agent Works
Step-by-step operational flow showing how this agent processes tasks end-to-end.
1
Step 1
**Risk Score** (0 to 100) is computed from three components:
2
Step 2
Blast radius (0 to 40): Based on action type; ranges from 5 (single agent config) to 35 (pause campaign)
3
Step 3
Reversibility (0 to 30): Lower for easily reversible actions; higher for actions with lasting effects
4
Step 4
Data confidence penalty (0 to 30): Higher penalty when impact estimates have low confidence
5
Step 5
**Confidence Score** (0 to 100) is computed from three components:
6
Step 6
Data quality (0 to 40): Based on root cause analysis confidence
7
Step 7
Pattern match (0 to 30): Based on similarity to past cases found in vector memory
8
Step 8
Historical success (0 to 30): Based on past success rate of the recommended action
9
Step 9
**Risk Tier** is assigned: safe (<25), moderate (25 to 49), risky (50 to 74), dangerous (75+)
10
Step 10
**Auto-execution gate**: Resolves thresholds from governance config (defaults: risk < 40, confidence > 70)
11
Step 11
Actions meeting both thresholds are flagged for auto-execution; others are flagged as "propose only"
Execution Modes
Trigger: event (invoked during the autonomous decision pipeline)
Data
Inputs & Outputs
What data this agent consumes and the artifacts or actions it produces.
Input Data
- `ActionRecommendation` from the Action Planner Agent
- `ImpactEstimate` from the Impact Estimator Agent
- `RootCauseResult` from the Root Cause Agent
Deliverables
- `RiskEvaluation` containing:
- Risk score (0 to 100)
- Confidence score (0 to 100)
- Risk tier (safe, moderate, risky, dangerous)
- Auto-executable flag (true/false)
- Reasoning chain explaining each scoring component
- Score breakdown by category
Core Tasks
- Strategic Intelligence
Integrations
Systems Connected
Internal systems, APIs, and tools this agent integrates with.
Tools & APIs
Receives inputs from **Action Planner Agent**, **Impact Estimator Agent**, and **Root Cause Agent**Reads governance thresholds from **GovernanceResolutionService**Output determines whether the **Execution Agent** auto-executes or the decision enters the proposal queueScoring breakdown is recorded in the **IntelligenceDecision** for audit
Specifications
Agent Specs
Technical specifications, requirements, and deployment details.
Related
Related Agents
Other agents in the same department or industry.
Enterprise AI
Ready to deploy this agent?
Schedule a walkthrough with our team to see how this agent integrates with your workflows.