explore

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly clones and analyzes arbitrary public GitHub repositories (SKILL.md workflow step "自动克隆 GitHub 仓库" and scripts/clone.sh which runs gh repo clone), so it ingests untrusted, user-generated third‑party content that the agent reads and uses for analysis and decision-making.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly runs gh repo clone to fetch arbitrary GitHub repositories at runtime (e.g., the owner/repo git URL passed to "gh repo clone"), and those fetched repository files are ingested for the agent's analysis so external repo content can directly control the agent's prompts/context.