dbdiagram
Pass
Audited by Gen Agent Trust Hub on Nov 22, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists entirely of markdown files containing technical documentation and agent instructions, with no executable logic.
- [INDIRECT_PROMPT_INJECTION]: While the skill ingests untrusted data in the form of DDL or database designs, it includes clear boundary instructions to mitigate injection risks.
- Ingestion points: Processes external DDL and design structures as primary inputs (SKILL.md).
- Boundary markers: Explicitly instructs the agent to treat source content as evidence rather than instructions and to ignore any credentials (SKILL.md).
- Capability inventory: Restricted to writing .drawio XML files based on structural data.
- Sanitization: Requires the agent to validate XML well-formedness and internal consistency of cell identifiers.
- [NO_CODE]: No Python, Node.js, or shell scripts are present in the skill directory.
Audit Metadata