The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified. The skill correctly limits its actions to the 'gh' CLI tool as specified in its configuration.
[COMMAND_EXECUTION]: The skill uses dynamic context injection (!) in SKILL.md to retrieve the current user and list of open maps. These commands are benign and limited to the 'gh' tool for context-gathering.
[PROMPT_INJECTION]: The skill processes data from GitHub issues, representing an indirect prompt injection surface. Ingestion points: Data enters via gh issue view and gh issue list (tracker-mechanics.md). Boundary markers: None are defined to isolate untrusted issue content. Capability inventory: The skill can execute gh commands and invoke other platform skills (SKILL.md). Sanitization: No escaping or validation is applied to external issue content.

wayfind