The Agent Skills Directory

[SAFE]: The skill acts as a set of structured instructions for an AI agent to generate social media copy. It provides specific platform constraints (character limits, hashtag strategies, hook placement) and includes clear examples for each target platform.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an ingestion surface for untrusted data by allowing users to provide file paths or URLs to existing content (blogs, newsletters, etc.) for repurposing. This is an inherent risk factor for skills processing external data, but it is necessary for the skill's primary function.
Ingestion points: Topic or source content (file path or URL) provided by the user in the input phase.
Boundary markers: The skill does not explicitly define delimiters or instructions to ignore embedded prompts within the source content.
Capability inventory: The skill is restricted to Read, Write, Glob, and Grep tools. It writes generated posts to the .jez/artifacts/ directory.
Sanitization: The skill instructions focus on extracting key messages but do not specify filtering or escaping techniques for the ingested content.
[COMMAND_EXECUTION]: The skill uses the Write tool to store generated artifacts in the .jez/artifacts/ directory. This is consistent with the vendor's (jezweb) workspace structure and represents legitimate file management functionality.
[EXTERNAL_DOWNLOADS]: A reference to Cloudflare (#CloudflareWorkers) appears in a LinkedIn post example. Cloudflare is a well-known service, and the mention is part of a benign educational example rather than a download or execution pattern.

social-media-posts