jp-prd

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.65). High likelihood: the required runtime workflow ingests user-provided free text and file contents from the target codebase/workspace (e.g., “Read the target codebase and environment” and “Ask for verbal context and any existing inputs they want you to read”), which can include outsider-authored text such as third-party repo files or forwarded/downloaded documents, and that extracted prose is then used as LLM context for PRD generation.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill calls doc_fetch_tool/explore_product at runtime to fetch Juspay documentation from the allowed domains (*.juspay.io, *.juspay.in, dth95m2xtyv8v.cloudfront.net), and those fetched pages are explicitly required to ground PRD facts and drive prompts, so this is a runtime external dependency that controls agent instructions.