The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses local scripts (e.g., skf-atomic-write.py) and standard utilities (e.g., npx skill-check) to handle file management, atomic commits, and output validation. These executions are part of the documented Stack Skill Forge (SKF) framework and are scoped to the local project environment.
[DATA_EXFILTRATION]: The skill reads project source code and manifests to extract architectural metadata and usage patterns. No network operations were found that send data to external or untrusted domains.
[PROMPT_INJECTION]: The instructions implement a strict, multi-step workflow with 'Zero hallucination' mandates and sequential execution rules. No patterns attempting to bypass safety filters or override the agent's core instructions were detected.
[INDIRECT_PROMPT_INJECTION]: The skill processes project-specific data like source code and architecture documents. Although it processes untrusted inputs, it extracts structural metadata (signatures, counts, patterns) rather than executing content, significantly limiting the surface for indirect injection attacks.
[EXTERNAL_DOWNLOADS]: The skill utilizes npx for validation tasks. The use of the --no-install flag in the validation steps minimizes the risk of downloading and executing unverified remote packages.

skf-create-stack-skill