create-specification
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input to generate documents, creating a potential surface for prompt injection if the resulting files are later consumed by an LLM.
- Ingestion points: The
${input:SpecPurpose}variable inSKILL.mdcaptures user input. - Boundary markers: The instructions do not use specific delimiters or escaping to isolate the user-provided input from the structural markdown.
- Capability inventory: The skill instructions include writing files to the
/spec/directory. - Sanitization: The skill does not perform any validation or sanitization on the input before incorporating it into the document generation process.
Audit Metadata