The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a local utility script scripts/bootstrap-experience.ts used to initialize the project's directory structure and experience layer. This script is executed locally to create and manage the .agents/experience/ workspace and does not perform any remote operations or unauthorized modifications.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from local research artifacts and user-provided briefs. While this creates a potential surface for indirect prompt injection, the skill mitigates this through its 'Pre-Dispatch Protocol'. This protocol includes a 'Warm Start' summary that requires the user to review and confirm all gathered context before sub-agents are spawned, ensuring human-in-the-loop validation of input data.
[SAFE]: The skill implements significant defensive measures including a dedicated Voice Agent to remove AI-generated patterns ('slop') and a Critic Agent that enforces strict quality rubrics, falsifiability tests, and competitor swap tests. No obfuscation, data exfiltration, or persistence mechanisms were found.

copywriting