pythia-explore

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill relies on a command-line utility named pythia to execute database exploration tasks such as listing objects, reading source code, and running SQL queries as documented in SKILL.md.\n- [INDIRECT_PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by ingesting and processing content from an external Oracle database.\n
  • Ingestion points: The agent retrieves potentially untrusted data through commands like pythia src (for PL/SQL source), pythia grep (for code search), and pythia sql (for arbitrary queries) in SKILL.md.\n
  • Boundary markers: The skill does not define specific delimiters or instructions to ensure the agent ignores instructions that might be embedded in the retrieved database content.\n
  • Capability inventory: The agent is empowered to read system-wide source code and execute SQL queries within the connected schema.\n
  • Sanitization: No sanitization or validation mechanisms are mentioned to process the database output before it is included in the agent context.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 04:12 AM
Security Audit — agent-trust-hub — pythia-explore