agent-api-designer
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill uses the
$ARGUMENTSplaceholder to ingest external input without implementing clear boundary markers or instructions to treat the data as untrusted content. - Ingestion points: User-provided input is injected via
$ARGUMENTSat the end ofSKILL.md. - Boundary markers: There are no delimiters (such as XML tags or unique markers) surrounding the input to isolate it from the core instructions.
- Capability inventory: As a general-purpose subagent, the skill is intended to generate code snippets, OpenAPI specifications, and documentation which may be written to the filesystem or executed by a parent agent.
- Sanitization: No input validation or sanitization logic is present to prevent embedded instructions from overriding the agent's behavior.
Audit Metadata