The Agent Skills Directory

[SAFE]: The skill serves as a documentation router and reference set for the Claude Agent SDK. Analysis of the included instructions and scripts found no malicious patterns, obfuscation, or unauthorized data access.- [EXTERNAL_DOWNLOADS]: The scripts/check-updates.py script performs HTTP requests to official registries (PyPI and npm) to retrieve version metadata. These operations target well-known, trusted domains for the purpose of skill maintenance.- [COMMAND_EXECUTION]: The skill documents the Bash tool provided by the Claude Agent SDK. This is a legitimate feature of the library, and the documentation includes a dedicated 'Secure Deployment' section advising on isolation techniques such as Docker hardening and gVisor to mitigate risks.

claude-agent-sdk