alterlab-research-lookup

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly sends queries to external search endpoints (Perplexity sonar-pro-search via OpenRouter at https://openrouter.ai/api/v1/chat/completions and the Parallel Chat API at https://api.parallel.ai), ingests their returned web citations/URLs/DOIs (see research_lookup.py and SKILL.md), and uses that untrusted, public-web content to build and drive research reports and citation-based decisions—exposing the agent to indirect prompt injection from arbitrary third-party pages.