skrapi

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

• Third-party content exposure detected (high risk: 0.75). The required workflow instructs the agent to inspect the real project files at runtime (e.g., package.json, config files, and directory tree) and then write/pull their contents into generated Markdown like ARCHITECTURE.md/PROMPT.md, so any outsider-authored text in the target repo (e.g., code/config authored by others) can be ingested into the LLM context indirectly during analysis.