The Agent Skills Directory

[GENERAL SAFETY] (SAFE): The skill consists entirely of instructional markdown and reference files. There are no scripts (.py, .js, .sh), no binary executables, and no configuration files that execute commands.
[REMOTE_CODE_EXECUTION] (SAFE): No remote code execution patterns were detected. There are no curl, wget, or dynamic execution commands present.
[EXTERNAL_DOWNLOADS] (SAFE): The skill does not perform any external downloads. All referenced materials (e.g., references/copy-frameworks.md) are local to the skill package.
[DATA_EXFILTRATION] (SAFE): There are no network-capable commands or functions that could facilitate data exfiltration.
[CREDENTIALS_UNSAFE] (SAFE): No hardcoded API keys, tokens, or sensitive credentials were found in any of the files.
[PROMPT_INJECTION] (SAFE): While the skill contains instructional language (e.g., 'You are an expert conversion copywriter'), it does not contain adversarial bypass attempts or instructions to ignore safety filters. It follows standard agent persona framing.
[INDIRECT_PROMPT_INJECTION] (LOW): The skill reads a local file (.claude/product-marketing-context.md) if it exists. While this is an ingestion point for external data, the risk is negligible as it is a local configuration file intended for context, and the skill lacks any capabilities (like shell execution or network access) that could be exploited via injection.

copywriting