Production-ready
Problem Discovery Agent
Scans business metrics and system state for anomalies, conversion drops, error spikes, and agent failures. Produces a deduplicated list of detected problems for the autonomous intelligence engine to investigate and resol
Intelligence & Analytics department for Colaberry Enterprise agents
Built by Colaberry
About the Agent
What this agent does, the challenges it addresses, and where it delivers value.
Scans business metrics and system state for anomalies, conversion drops, error spikes, and agent failures. Produces a deduplicated list of detected problems for the autonomous intelligence engine to investigate and resolve.
Challenges This Agent Addresses
- 1**IT Operations**: Detects agents stuck in error state and escalates based on failure persistence
- 2**Marketing**: Identifies sudden drops in lead generation that may indicate campaign or funnel issues
- 3**Platform Engineering**: Catches error rate spikes early before they cascade into system-wide failures
How the Agent Works
Step-by-step operational flow showing how this agent processes tasks end-to-end.
Step 1
Runs three detection checks in parallel:
Step 2
**Agent failures**: Finds enabled agents in "error" status; severity escalates with error count
Step 3
**Conversion drops**: Compares 48-hour lead count against the 7-day daily average; flags drops exceeding 40%
Step 4
**Error spikes**: Compares last-hour error count against 24-hour hourly average; flags spikes exceeding 3x the norm
Step 5
Collects all detected problems from the parallel checks
Step 6
Deduplicates against vector memory by searching for similar recent entries (within 2 hours, similarity above 90%)
Step 7
Returns only genuinely new problems for downstream investigation
Execution Modes
Inputs & Outputs
What data this agent consumes and the artifacts or actions it produces.
Input Data
- Agent status records (from `AiAgent` table)
- Lead generation metrics (from `leads` table)
- System process error counts (from `system_processes` table)
- Vector memory for deduplication against recently reported issues
Deliverables
- List of `DetectedProblem` objects, each containing:
- Problem type (kpi_anomaly, conversion_drop, error_spike, engagement_decline, agent_failure, pipeline_bottleneck)
- Severity (low, medium, high, critical)
- Entity type and ID (when applicable)
- Description of the issue
- Relevant metrics
- Detection timestamp
Core Tasks
- Strategic Intelligence
Systems Connected
Internal systems, APIs, and tools this agent integrates with.
Tools & APIs
Agent Specs
Technical specifications, requirements, and deployment details.
Related Agents
Other agents in the same department or industry.
Ready to deploy this agent?
Schedule a walkthrough with our team to see how this agent integrates with your workflows.