The Agent Skills Directory

[SAFE]: The skill instructions and reference materials are purely text-based and do not include any scripts, tools, or dangerous functionality.\n- [PROMPT_INJECTION]: No instructions were found that attempt to subvert safety filters, extract system prompts, or hijack agent behavior.\n- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill references brand-specific context files (e.g., .agents/product-marketing.md), which is a benign use of local project context and does not involve exfiltrating sensitive credentials or system files.\n- [UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: There are no remote scripts, package installations, or command executions defined in the skill or its references.\n- [OBFUSCATION]: The analysis found no hidden text, encoded commands, or obfuscated URLs within the markdown or JSON files.\n- [INDIRECT_PROMPT_INJECTION]: While the skill processes external user text for editing, it lacks the necessary capabilities (such as network access or file system writing) to be exploited via injection.

copy-editing