context7

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

• Third-party content exposure detected (high risk: 0.80). This skill calls Context7's public API (e.g., https://context7.com/api/v2/libs/search and /api/v2/context) to fetch live library documentation and code snippets from external/public sources that the agent reads and uses to form responses, exposing it to untrusted third-party content that could carry indirect prompt injections.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly performs runtime requests to https://context7.com/api/v2/... to retrieve documentation that is injected into the agent's responses (thus directly controlling prompts) and the skill depends on that external content.