monitoring-database-health
Warn
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: MEDIUMPERSISTENCEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [PERSISTENCE]: The skill instructions direct the agent to 'Compile all health checks into a single monitoring script that runs via cron every 60 seconds'. This establishes a persistence mechanism on the host system, allowing code to execute periodically outside of the agent interactive session.
- [DYNAMIC_EXECUTION]: The skill requires the agent to generate and potentially execute new shell or Python scripts to perform health checks and automated remediation tasks, such as restarting database services.
- [INDIRECT_PROMPT_INJECTION]: The skill operates by ingesting data from external database CLI tools, which represents an attack surface for indirect prompt injection. 1. Ingestion points: Output from psql, mysql, and mongosh commands (SKILL.md). 2. Boundary markers: Absent. 3. Capability inventory: Write and Edit permissions for script creation, plus Bash access to database CLI tools (SKILL.md frontmatter). 4. Sanitization: Absent; no validation of database-returned data is specified.
Audit Metadata