hypothesis-generation

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

• Third-party content exposure detected (high risk: 0.75). The workflow explicitly requires runtime literature retrieval via “WebFetch with PubMed URLs” and “WebSearch” (public web content), whose fetched page text can be ingested into the LLM context for synthesis, creating an indirect prompt-injection exposure path.

MEDIUM W021: Hidden or invisible Unicode characters detected (potential obfuscation or prompt injection).

• Hidden Unicode characters detected (1 type(s) found)