keyword-research

Pass

Audited by Gen Agent Trust Hub on Oct 6, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill configures a connection to the Ubersuggest MCP server at https://ubersuggest-mcp.neilpatelapi.com/mcp as defined in agents/openai.yaml. This targets the vendor's official API infrastructure and is a functional requirement for retrieving SEO metrics.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes search suggestions and metrics from external sources, which constitutes a potential ingestion surface. 1. Ingestion points: User input from $ARGUMENTS and keyword data retrieved from tools such as google_suggestions and keyword_suggestions. 2. Boundary markers: The instructions do not specify boundary markers or instructions to ignore embedded commands. 3. Capability inventory: The skill is restricted to calling SEO-specific tools and generating data tables, with no access to sensitive system resources or arbitrary network destinations. 4. Sanitization: No explicit sanitization of tool output is defined.
  • [SAFE]: No malicious patterns, such as hardcoded credentials, prompt injection, or obfuscated code, were detected. The skill's logic is consistent with its stated purpose of SEO research.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 6, 2026, 07:49 PM