Skip to main content

​
How Issue Detection Works

Sentrial analyzes every session automatically to identify problems. No manual tagging required.

Automatic Scanning

Every session is analyzed in real-time as events stream in.

AI Analysis

LLM evaluates sessions for failures, confusion, errors, and inefficiencies.

Smart Grouping

Similar issues are grouped together so you see patterns, not noise.

​
What We Detect

Sentrial identifies a wide range of issues across your agent sessions:
IssueDescriptionSeverity
Task FailuresAgent failed to complete the user’s requestHigh
Tool ErrorsA tool call failed, returned unexpected results, or timed outHigh
HallucinationsAgent made claims not supported by context or toolsHigh
User FrustrationSigns of user dissatisfaction, repeated questions, or negative sentimentMedium
Confusion LoopsAgent appears stuck, repeating actions or asking the same questionsMedium
Slow ResponsesSession took significantly longer than expectedLow
High CostSession used more tokens/cost than typicalLow

​
AI-Powered Diagnosis

When you click on an issue, Sentrial provides:
  • Root Cause Analysis — Why did this happen? What triggered it?
  • Affected Sessions — All sessions impacted by this issue
  • Recommended Fix — Specific suggestions to resolve the issue
  • Code Context — Links to relevant code if GitHub is connected

​
The Signals Page

The Signals page in your dashboard is your command center for issues:

Issue List

All detected issues sorted by severity and frequency.

Trend Charts

See how issue counts change over time.

Filters

Filter by agent, time range, severity, or issue type.

Quick Actions

Jump to diagnosis, affected sessions, or code fixes.

​
Fixing Issues in Code

Code Integration — Connect your GitHub repository and fix issues directly from the diagnosis page. Chat with AI about the issue, get code suggestions, and create PRs — all in one flow. Learn about Code Integration →

​
Best Practices

Always call complete_session() with success=True/False so we can identify failures.
When success=False, include failure_reason for better diagnosis.
Group sessions by agent_name to see issues per agent type.
Include user_id to identify user-specific issues and patterns.
# Good: Complete session with full context
client.complete_session(
    session_id=session_id,
    success=False,
    failure_reason="Tool 'search_kb' returned empty results for query",
    estimated_cost=0.023
)

# This gives Sentrial the context to:
# - Identify this as a tool failure issue
# - Group similar empty-result failures
# - Suggest fixes (e.g., improve search, add fallback)

​
Next Steps

Code Integration

Fix issues directly in your codebase.

Sessions & Tracking

Learn how to track sessions properly.