The Agent Skills Directory

[SAFE]: No security issues were detected. The skill focuses on providing a framework for creating project plans and uses standard file-reading tools (Read, Glob, Grep) for context.
[PROMPT_INJECTION]: The skill uses natural instructional language to guide planning activities. No bypass, override, or jailbreak patterns were identified.
[INDIRECT_PROMPT_INJECTION]: The skill reads project files to generate task lists, which is a standard operational surface. 1. Ingestion points: Project files accessed via Read, Glob, and Grep tools. 2. Boundary markers: None specified in the instructions. 3. Capability inventory: File reading and pattern matching; no code execution or network operations are included in the skill definition. 4. Sanitization: None. This surface is considered safe given the skill's limited capabilities.
[REMOTE_CODE_EXECUTION]: While the skill mentions project-specific scripts like 'security_scan.py', these are documented as examples of content to include in a plan rather than code to be fetched or executed by the skill itself.

plan-writing