The Agent Skills Directory

[SAFE]: No malicious patterns such as prompt injection, data exfiltration, or obfuscation were detected. The skill follows established safety guidelines and focuses entirely on its stated purpose of assisting with copywriting.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it processes untrusted user-supplied briefs and article drafts.
Ingestion points: User input regarding article topics, audiences, and draft content (SKILL.md).
Boundary markers: The skill incorporates human-in-the-loop validation by confirming the brief before generation and using the AskUserQuestion tool to let the user select the final output.
Capability inventory: The skill has access to file manipulation tools (Read, Edit, Write, Glob, Grep) and the Agent invocation tool, which are standard for the development harness.
Sanitization: While no explicit programmatic sanitization is defined, the high level of user oversight and the specific formatting requirements for outputs minimize the exploitability of this surface.

copywriting-hooks