pythia-apply
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a comprehensive safety framework for database operations, requiring verbatim relay of impact analysis and diffs to the user to ensure manual oversight of all changes.
- [SAFE]: The instructions explicitly prohibit autonomous approval or bypassing established security gates (such as using --yes or minting approval tokens), ensuring the agent cannot execute writes without manual developer intervention.
- [SAFE]: No evidence of credential exposure, data exfiltration, or malicious persistence mechanisms was found within the skill content.
- [INDIRECT_PROMPT_INJECTION]: The skill provides an interface for processing database CLI outputs which represents an ingestion surface. This is mitigated by the following controls documented in SKILL.md: 1) Ingestion point: Output from the pythia apply command. 2) Boundary markers: Mandatory instructions to relay all diffs and warnings verbatim. 3) Capability inventory: Shell command execution via the pythia tool suite. 4) Sanitization: Mandatory human-in-the-loop review and approval of the exact raw output before any write operation is finalized.
Audit Metadata