The Agent Skills Directory

[SAFE]: The skill follows security best practices for secret management by instructing the agent to load the ZHIPU_API_KEY from an environment variable rather than hardcoding it. All external references and API endpoints target official Zhipu AI service domains (open.bigmodel.cn and docs.bigmodel.cn).
[INDIRECT_PROMPT_INJECTION]: The skill defines a mechanism for the agent to process user-provided messages and send them to an external LLM API, which is a standard attack surface for indirect prompt injection.
Ingestion points: The messages field in the API request body.
Boundary markers: None defined.
Capability inventory: Network POST requests to Zhipu AI endpoints.
Sanitization: No sanitization or input validation logic is described.
[EXTERNAL_DOWNLOADS]: The Python SDK code snippet uses from zai import ZhipuAI. Users should be aware that the official Zhipu AI SDK package name on PyPI is zhipuai; the use of zai may be a documentation error or refer to an unofficial library, though no installation commands are present in the skill to automate its deployment.

zhipu-text