The Agent Skills Directory

[SAFE]: The skill core functionality is restricted to reading and analyzing markdown files to provide feedback in the chat. It does not perform any high-risk actions such as network communication, privilege escalation, or persistence.
[SAFE]: The skill accesses specific local files including draft.md, outline.md, and keyword-data.md. This file access is limited to reading content for evaluation purposes and is well-defined within the project directory structure.
[PROMPT_INJECTION]: The skill processes untrusted user data from blog drafts which presents a potential injection surface. 1. Ingestion points: draft.md, outline.md, and keyword-data.md. 2. Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present. 3. Capability inventory: Restricted to text processing and conversation output; no file writing, network access, or command execution capabilities were found. 4. Sanitization: No evidence of input sanitization or filtering was detected. Given the lack of dangerous capabilities, this surface poses no significant risk.

seo-review