The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill utilizes the WebFetch tool to download content from external URLs provided by the user for SEO analysis. This is a standard functional component for analyzing live web pages.
[COMMAND_EXECUTION]: The skill executes local Python scripts (skills/blog-google/scripts/run.py) to perform Google authentication checks and run PageSpeed tests. This involves subprocess execution within the agent environment.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads and processes external content (via WebFetch or file system access) which is then used to generate a report.
Ingestion points: Step 1 extracts frontmatter, headings, links, and body content from target files or fetched URLs.
Boundary markers: None present; the skill does not wrap processed content in delimiters or include instructions to ignore embedded commands.
Capability inventory: The skill has access to subprocess execution (python3), network operations (WebFetch), and file system scanning (Grep, Glob).
Sanitization: No evidence of input validation or escaping of the ingested content is found in the instructions.
[DATA_EXFILTRATION]: The skill uses Grep and Glob to scan the local project environment to verify internal and bidirectional links, which involves reading multiple files within the workspace.

blog-seo-check