seo-competitive-intel
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The
competitor-content/SKILL.mdmodule utilizesweb_fetchto retrieve HTML and sitemaps from competitor websites for analysis. This is a core feature for monitoring competitor updates but involves interacting with external, untrusted web servers. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8).
- Ingestion points: The system fetches HTML content (titles, H1-H4 headers, and body text) from competitor websites using
web_fetchin thecompetitor-contentsub-module. - Boundary markers: No explicit delimiters or 'ignore' instructions are used when processing the scraped HTML content.
- Capability inventory: The skill generates 'Action Plan' recommendations and produces a
.jsxfile based on the analysis of this external data. - Sanitization: There is no mention of sanitizing or filtering the scraped content for hidden instructions before it is processed by the agent to generate advice.
- [COMMAND_EXECUTION]: The
monthly-report/SKILL.mdmodule dynamically generates a.jsxfile (React component) containing the monthly SEO data and action plans. This represents the generation of executable UI code based on aggregated data. - [CREDENTIALS_UNSAFE]: The skill requires the
SEMRUSH_API_KEYenvironment variable and OAuth access to Google services. While no secrets are hardcoded, users should ensure their environment handles these credentials securely as per best practices.
Audit Metadata