alibabacloud-find-skills

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

• Third-party content exposure detected (medium risk: 0.65). SKILL.md’s required runtime workflow calls the AgentExplorer HTTP API via curl and can fetch outsider-authored skill markdown content (from the public skills catalog) which is readable free text that may be inserted into the agent’s LLM context.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill makes runtime HTTP requests to https://agentexplorer.aliyuncs.com (e.g. GET /openapi/for-agent/skills and /openapi/for-agent/skills/{skillName}) to fetch skill markdown that the agent may follow, and the documented installation step (npx skills add) fetches and executes remote code — so these external endpoints can directly control prompts or execute code at runtime.