reply-maker

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill instructs building and sending curl commands that include an Authorization: Bearer {XAI_API_KEY} header (and requires the injected XAI_API_KEY), which forces the agent to place a secret into command output/requests and creates an exfiltration risk.

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

• Third-party content exposure detected (medium risk: 0.30). SKILL.md’s Mode A “Gather candidate tweets” builds an X.AI request that includes “full tweet text” fetched from X via the x_search tool (i.e., outsider-authored tweets) and then parses those texts for reply generation.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill mandatorily performs a runtime POST to https://api.x.ai/v1/responses using the injected XAI_API_KEY and consumes the returned model output to assemble candidate tweets and drive reply generation, so that external endpoint directly controls prompts and is a required runtime dependency.