seo-wordpress-manager
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it processes untrusted content from a WordPress site and uses it to generate suggestions for the user.\n
- Ingestion points: WordPress post content, excerpts, and current SEO fields are fetched using GraphQL in
scripts/analyze_seo.pyand passed to the agent for analysis.\n - Boundary markers: The instructions generated for the agent in
scripts/analyze_seo.pydo not include strict delimiters or warnings to ignore instructions that might be embedded within the post content being optimized.\n - Capability inventory: The skill is capable of performing write operations (mutations) to the WordPress database via the
update_post_seomethod inscripts/yoast_batch_updater.py.\n - Sanitization: Although basic HTML stripping is performed on post content, there is no logic to detect or neutralize natural language instructions (prompt injections) that could influence the agent's output.
Audit Metadata