case-study-writing

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly instructs the agent to run web searches and fetch public content (e.g., "infsh app run tavily/search-assistant" in Quick Start and "infsh app run exa/search" / "exa/answer" in Research Support) to gather industry benchmarks and statistics that the agent is expected to read and use in crafting the case study, exposing it to untrusted third-party content that can influence outputs.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly calls the inference.sh CLI and remote apps at runtime (e.g., infsh app run tavily/search-assistant and infsh/python-executor), meaning https://inference.sh is a required external runtime dependency that runs remote apps (including a Python executor) which can execute code and influence agent behavior.