claude-agent-sdk

Pass

Audited by Gen Agent Trust Hub on Apr 22, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Documents installation of the official claude-agent-sdk (Python) and @anthropic-ai/claude-agent-sdk (TypeScript) packages from reputable registries (PyPI and NPM).
  • [COMMAND_EXECUTION]: Describes the Bash tool which provides terminal access. The documentation includes security patterns for restricting command execution via hooks and permission modes.
  • [DATA_EXFILTRATION]: Details tools for file system access (Read, Grep) and web access (WebSearch, WebFetch). It provides guidance on using hooks to prevent access to sensitive files like .env.
  • [PROMPT_INJECTION]: Explains the bypassPermissions and acceptEdits modes which allow for autonomous tool execution. The documentation provides clear warnings that these modes should only be used within secured, containerized environments.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 22, 2026, 12:35 PM
Security Audit — agent-trust-hub — claude-agent-sdk