wiki-enrich

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

• Third-party content exposure detected (high risk: 0.95). Yes—Step 2.3 fetches outsider-authored free text from public websites at runtime (e.g., WebFetch https://alphaxiv.org/... or the arXiv API via curl), which is then passed as $SOURCE_TEXT into the agent LLM for Step 2.4 section generation.

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

• Potentially malicious external URL detected (high risk: 0.90). This skill fetches external documents at runtime and injects them as $SOURCE_TEXT that directly steers the LLM's generation (controlling prompts), e.g. https://alphaxiv.org/overview/<arxiv_id>.md, https://alphaxiv.org/abs/<arxiv_id>.md, and http://export.arxiv.org/api/query?id_list=<arxiv_id>.