claude-code-sdk
Pass
Audited by Gen Agent Trust Hub on Mar 9, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: This skill serves exclusively as a technical reference library for the Claude Code SDK, covering plugins, hooks, MCP, and subagents. All included files are Markdown-based documentation; there are no executable scripts, shell commands, or binary assets provided within the skill package.
- [SAFE]: The documentation content is security-aware and provides proactive guidance. It explicitly instructs developers to avoid exposing sensitive files such as .env, .git, or private keys, and provides warnings about the risks of using high-privilege configuration flags like --dangerously-skip-permissions.
- [NO_CODE]: The skill does not bundle any functional code, scripts, or logic, functioning strictly as a knowledge base to be queried by the agent for understanding the Claude Code platform internals.
Audit Metadata