seo-coach
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions for the user to connect to the HarborRank remote MCP server at
https://app.harborrank.com/mcp. This is a functional requirement for the skill to access SEO data and originates from the service provider's domain. - [INDIRECT_PROMPT_INJECTION]: The skill has a potential attack surface for indirect prompt injection as it is designed to ingest and process data from external sources.
- Ingestion points: The agent is instructed to read content from the user's website, competitor web pages, and local files (such as GSC CSV exports) to ground its SEO advice.
- Boundary markers: The instructions conceptually distinguish between 'live SEO data', 'web evidence', and 'coaching judgment', but do not define strict technical delimiters (like XML tags or specific markers) to wrap external content.
- Capability inventory: The agent has the ability to execute HarborRank MCP tools (keyword research, domain overviews), read files, crawl websites, and write to a
README.mdfile in the project directory. - Sanitization: There are no explicit instructions for the agent to sanitize or escape content retrieved from external websites before processing it or incorporating it into the workspace.
Audit Metadata