serilog
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides comprehensive guidance on structured logging, which is a security best practice. It explicitly warns against logging sensitive information and demonstrates how to use message templates to prevent log injection vulnerabilities.
- [INDIRECT_PROMPT_INJECTION]: The skill documents how to log data received from external sources, identifying the inherent attack surface of logging systems while providing mitigation strategies.
- Ingestion points: Data enters the logging context through service methods (SKILL.md), plugin initialization (SKILL.md), and telemetry hubs (references/workflows.md).
- Boundary markers: The skill enforces the use of Serilog's structured message templates (e.g., 'Tank {TankId} updated'), which effectively delimits trusted template strings from untrusted property values.
- Capability inventory: The skill utilizes standard tools like 'Read', 'Edit', 'Write', and 'Grep' to manage and query log files stored in the 'logs/' directory.
- Sanitization: The 'Anti-Patterns' section in references/patterns.md specifically provides guidance on redacting sensitive values like passwords before they reach the logs.
Audit Metadata