The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted data from multiple sources including user-provided URLs via webReader, remote images, and local HTML files, creating a surface for indirect prompt injection. 1. Ingestion points: Input parameters, remote URLs, and local HTML file paths. 2. Boundary markers: Absent. The instructions do not define delimiters to isolate external content. 3. Capability inventory: The skill can write files to the docs/product/ directory. 4. Sanitization: Absent. There is no mention of validating or escaping external data before it is incorporated into the generated PRD.
[COMMAND_EXECUTION]: The skill dynamically creates interactive HTML prototypes based on the generated requirements. Generating executable code from untrusted input is a potential attack surface for script injection in the generated artifacts.
[EXTERNAL_DOWNLOADS]: The skill interacts with external URLs and images to collect context for requirements, which involves network operations targeting user-specified locations.

prd-gen