agent-tool-builder
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill describes building tools that ingest data from external sources, which creates a potential attack surface.
- Ingestion points: Defined in tool patterns such as 'search_web' and API integrations in references/detailed-guide.md.
- Boundary markers: The documentation advises users to treat all tool results as untrusted content.
- Capability inventory: Uses standard SDK runners for tool execution.
- Sanitization: The guide specifically mandates strict validation and SQL parameterization.
- [SAFE]: No malicious obfuscation, credentials harvesting, or remote code execution patterns were detected.
Audit Metadata