The Agent Skills Directory

[SAFE]: The skill consists entirely of instructional Markdown content and templates for organizational planning. No malicious patterns, obfuscation, or unauthorized command executions were found in any of the skill files.\n- [PROMPT_INJECTION]: The workflow involves gathering and processing inputs from external sources such as product roadmaps and customer feedback (file SKILL.md). This ingestion surface lacks explicit boundary markers or sanitization, which is a risk factor for indirect prompt injection. However, given the primary purpose is document generation and the agent is restricted to standard file operations (Read, Write, Edit), the vulnerability is not actionable for significant harm.

technical-roadmap