The Agent Skills Directory

[DYNAMIC_EXECUTION]: The script scripts/audit-seo.mjs dynamically imports and executes configuration files (e.g., bascik.config.js, .mjs, .ts) from the project root using await import() with a computed filesystem path.
[DATA_EXFILTRATION]: The audit-seo.mjs script accesses and reads the .env file in the project root to extract the BASCIK_SITE_URL variable. Accessing sensitive file paths like .env is a data exposure risk.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes HTML content from the build output directory to generate SEO audits, which could be used as a vector for instructions if the build includes untrusted data.
Ingestion points: scripts/audit-seo.mjs reads all .html files found within the dist/ directory.
Boundary markers: The script does not use specific delimiters or warnings to separate site content from the auditor's instructions.
Capability inventory: The script has permissions to read local files and execute local JavaScript configuration files.
Sanitization: There is no sanitization of extracted HTML attributes (like titles or meta descriptions) before they are reported to the agent context.

bascik-seo