extended-thinking
Pass
Audited by Gen Agent Trust Hub on Aug 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves as an educational and configuration guide for enabling 'extended thinking' in Claude models. It provides Python and TypeScript code snippets demonstrating how to interact with the Anthropic API.
- [SAFE]: No executable code is provided that would run automatically on the user's system. All code examples are for integration purposes and do not contain hidden payloads or obfuscation.
- [SAFE]: The listed
allowed-tools(Bash, Read, Write, etc.) are standard platform capabilities and their inclusion in the frontmatter is a configuration limit, not a security finding. - [SAFE]: No hardcoded credentials, sensitive file access, or unauthorized network operations were detected. All API endpoints and documentation refer to official Anthropic services.
Audit Metadata