The Agent Skills Directory

[SAFE]: The skill implements a robust multi-agent orchestration architecture for generating high-quality copywriting. It separates concerns across multiple agents (Hook, Body, CTA, etc.) and uses a Critic agent to enforce quality gates.
[SAFE]: Utility scripts such as manifest-sync.ts, append-loop-result.ts, and scaffold-eval-loop.ts facilitate artifact management and evaluation loops. These scripts include explicit security measures, such as verifying that directories are not symbolic links and ensuring file operations remain within the project's designated loop folders.
[SAFE]: The allowed-tools configuration is appropriate for the skill's functionality, enabling reading, searching, and fetching web content for research purposes.
[SAFE]: No evidence of prompt injection was found. The agent instructions are focused on craft knowledge (e.g., PAS framework, V/F/U rubric) and do not attempt to bypass safety filters or override system behavior.
[SAFE]: No data exposure or exfiltration risks were identified. The skill does not access sensitive system files or credentials, and network operations are targeted at well-known services used as examples in the copy (e.g., Stripe, Vercel).

write-copy