wiki-adapter-prd

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

• Third-party content exposure detected (high risk: 0.85). The required workflow explicitly supports URL/foreign-document ingestion by running skillwiki fetch-guard <url> and then fetching the page body for “Raw capture (verbatim)” (raw/articles/<slug>.md), which is outsider-authored free text that can be transformed and fed into the LLM during “Generate typed-knowledge pages.”

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly runs runtime fetches for user-supplied URLs via the command "skillwiki fetch-guard " and then ingests that remote document into the agent's context (flagged token: ), meaning external content can directly influence prompts and outputs.