The Agent Skills Directory

[SAFE]: The skill implements secure authentication using the official alibabacloud-credentials SDK, adhering to the Alibaba Cloud default credential chain. It avoids hardcoded secrets and correctly instructs users to manage sensitive information via environment variables or secure configuration files.
[SAFE]: Network activity is restricted to the official STAROps API endpoint (starops.cn-beijing.aliyuncs.com), which is consistent with the skill's stated purpose and vendor identity (aliyun).
[INDIRECT_PROMPT_INJECTION]: The skill facilitates communication between the user and an AI diagnostic agent, which involves processing natural language input.
Ingestion points: Untrusted data enters the agent context through the --question command-line argument used in scripts/call_starops_agent.py.
Boundary markers: The script does not utilize explicit prompt delimiters for the question text, relying on the structured API request body.
Capability inventory: The skill is capable of performing network requests (POST /chat) to the STAROps backend to retrieve diagnostic information.
Sanitization: External input is properly sanitized and escaped using standard JSON serialization (json.dumps) before being transmitted in the API request body.

alibabacloud-starops-chat