seo-sxo
Pass
Audited by Gen Agent Trust Hub on Nov 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from target URLs and search engine results, creating a vulnerability surface for indirect prompt injection attacks.
- Ingestion points: Target URLs are fetched via
render_page.pyand parsed byparse_html.py. Search result data (snippets, People Also Ask questions, ad copy, and AI Overviews) are ingested via theWebSearchtool and potentiallyDataForSEOMCP tools. - Boundary markers: The skill includes a 'Security' section in
SKILL.mdthat explicitly instructs the agent to treat fetched content as untrusted data and ignore any instructions or commands found within that content. - Capability inventory: The skill has the capability to execute internal scripts (
render_page.py,parse_html.py) and perform network operations via search tools and external SEO APIs. - Sanitization: The skill utilizes a specific HTML parser to extract structured metadata rather than processing raw HTML, and it provides clear behavioral guardrails for the agent.
- [COMMAND_EXECUTION]: The skill instructions specify the execution of external scripts to perform its core functions.
- Evidence:
SKILL.mdreferences the execution ofrender_page.pyandparse_html.pyusing a command wrapper located in${CLAUDE_PLUGIN_ROOT}/scripts/claude-seo. These scripts appear to be local components of the skill's environment.
Audit Metadata