ptlam-creating-skill
Pass
Audited by Gen Agent Trust Hub on Aug 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill package consists exclusively of Markdown and YAML documentation. It does not contain executable code, scripts, or binary assets, and does not perform any automated network or file system operations beyond reading documentation.
- [EXTERNAL_DOWNLOADS]: Reference files include links to official platform documentation from Anthropic. These links are trusted and informational, serving as guidance for developers without triggering automated downloads or remote code execution.
- [PROMPT_INJECTION]: The skill functions as a meta-tool for processing user workflows into agent instructions. Ingestion points: Reads repository instructions and conversation history (SKILL.md). Boundary markers: Recommends using descriptive XML tags and headings to isolate instructions from user data (prompting-best-practices.md). Capability inventory: Instructions are limited to reading local files and generating markdown text. Sanitization: Encourages static verification against host-specific schemas to ensure structural validity.
- [COMMAND_EXECUTION]: The skill mentions running static validators, formatters, and drift checks as part of its verification process. These are standard development tasks targeting local repository files and do not include patterns for arbitrary command injection.
Audit Metadata