parallel-web

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

• Third-party content exposure detected (high risk: 0.85). Outsider free text from public web pages is ingested into the agent’s LLM context via the Parallel Chat API’s web research/synthesis path (ParallelSearch/ParallelDeepResearch calling client.chat.completions.create with the user objective, where the model incorporates retrieved page excerpts and returns them as content/sources.excerpts).