The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied data such as campaign names, search terms, and URLs, which are interpolated directly into the generated audit findings and fix lists.
Ingestion points: The run function in resources/search-audit.js accepts campaigns, searchTerms, and landingPages objects from the user configuration.
Boundary markers: No specific boundary markers or instructions to ignore embedded commands are present in the output generation logic.
Capability inventory: The skill is designed for use with agents (Claude Code, Cursor, Windsurf) that have file system access, shell execution, and network capabilities.
Sanitization: The script does not sanitize or escape string fields (e.g., c.name, t.term, p.url) before including them in the finding text, allowing potential injection of instructions into the agent's context.
[SAFE]: The skill operates entirely locally using a Node.js script to process data. It does not perform any network operations, external downloads, or remote code execution.
[SAFE]: No sensitive file access patterns, credential harvesting, or persistence mechanisms were detected in the skill scripts or configuration.

paid-search-account-audit