seo-cluster

Pass

Audited by Gen Agent Trust Hub on Nov 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data fetched via WebSearch and DataForSEO APIs (SERP titles, descriptions, and URLs). This content is used to generate content clusters and interactive maps. Maliciously crafted search results could attempt to influence the agent's behavior.
  • Ingestion points: Web search results and API responses integrated into the planning workflow in SKILL.md and references/serp-overlap-methodology.md.
  • Boundary markers: The skill includes a 'Security' section in SKILL.md explicitly instructing the agent to treat fetched content as untrusted data and never as instructions.
  • Capability inventory: Capability to write files (cluster-map.html, cluster-plan.json), execute local scripts, and invoke the blog-write skill.
  • Sanitization: The instructions mandate escaping characters (e.g., 0c for <) to prevent script injection in the generated HTML.
  • [COMMAND_EXECUTION]: The skill executes local scripts located in the plugin's root directory (scripts/claude-seo) to perform cost checks and render pages. While these are internal scripts, the execution of subprocesses based on user-triggered workflows is an area for caution.
  • [EXTERNAL_DOWNLOADS]: The skill fetches content from external URLs (via render_page.py) and third-party APIs (DataForSEO) to gather keyword and SERP data. This is a primary function of the skill but introduces external data into the execution environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 15, 2026, 01:51 PM