finding-your-unknowns
Pass
Audited by Gen Agent Trust Hub on Aug 16, 2026
Risk Level: SAFENO_CODEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists entirely of Markdown documentation and templates. No scripts, binaries, or executable code are included, minimizing the direct attack surface.
- [SAFE]: All external references point to the official blog of a well-known AI service provider, which is recognized as a reputable source for software development methodologies.
- [PROMPT_INJECTION]: The workflow instructs the agent to ingest external data, such as existing codebases or third-party libraries (e.g., the 'References' pattern in
references/prompt-library.md). This creates a surface for indirect prompt injection where malicious instructions embedded in the processed code could influence the agent. - Ingestion points: The
references/prompt-library.mdfile contains templates that instruct the agent to read and reimplement behavior from external source code paths. - Boundary markers: The templates do not include specific delimiters or instructions to ignore embedded commands within the analyzed code.
- Capability inventory: The skill does not provide its own scripts or tools; it operates using the agent's baseline capabilities.
- Sanitization: No sanitization or validation of the external content is prescribed in the workflow.
Audit Metadata