database-admin

Pass

Audited by Gen Agent Trust Hub on Aug 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill defines the use of the Bash tool to perform administrative database tasks such as pg_dump, pg_restore, and REINDEX. These commands are standard for the skill's purpose of database administration.
  • [DATA_EXFILTRATION]: Instructions are provided for automating database backups to S3 buckets and sending performance metrics to third-party monitoring services like Honeybadger and pganalyze. These represent standard industry practices for Rails application maintenance.
  • [PROMPT_INJECTION]: The skill processes live data from database engines which constitutes a surface for indirect prompt injection.
  • Ingestion points: Data enters the agent's context through SQL query results from monitoring views like pg_stat_statements and pg_stat_activity as seen in references/postgresql.md and SKILL.md.
  • Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat database results as untrusted data.
  • Capability inventory: The skill is configured with Bash, Read, Grep, and Glob tools, allowing for file system access and shell command execution based on processed data.
  • Sanitization: No sanitization or validation logic is defined for the metrics or logs retrieved from the database before they are interpreted by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 1, 2026, 08:54 PM
Security Audit — agent-trust-hub — database-admin