gpt-researcher

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

• Third-party content exposure detected (high risk: 1.00). The SKILL.md explicitly instructs the agent to run web searches and scrape public sources (e.g., Tavily search, Google/Bing retrievers, MAX_URLS_TO_SCRAPE) and to read saved results like /tmp/tavily_results_*.md as input for report synthesis, so it consumes untrusted third‑party web content that can influence subsequent actions.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly runs a Tavily search (via /tmp/gptr-env/bin/python ~/.claude/scripts/tavily_search.py) using the Tavily service (https://app.tavily.com) at runtime and then directs Claude to "Read /tmp/tavily_results_*.md"—meaning externally-fetched search results are injected into the model context and thus can directly control prompts, so this is a runtime external dependency that controls agent instructions.