plan-error-handling

Pass

Audited by Gen Agent Trust Hub on Jul 24, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill serves a clear, benign purpose as a senior reliability engineer agent, focusing on identifying silent failures and observability gaps in Sentry and Langfuse integration.
  • [SAFE]: Explicit guardrails are provided that restrict the agent to 'plan only' and 'no code or SDK edits,' ensuring no unauthorized modifications are performed without explicit user approval of the phases.
  • [SAFE]: The skill proactively addresses security concerns by instructing the agent to flag PII (Personally Identifiable Information) and secret leakage into capture systems, recommending redaction and scrubbing techniques (e.g., Sentry's beforeSend).
  • [SAFE]: No signs of prompt injection, obfuscation, credential exfiltration, or unauthorized network operations were found. The tool uses standard triggers and a structured reporting template.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 24, 2026, 11:19 AM
Security Audit — agent-trust-hub — plan-error-handling