seo-drift

Pass

Audited by Gen Agent Trust Hub on Sep 24, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a series of Python scripts—including fetch_page.py, parse_html.py, and drift_baseline.py—using a local runner located at ${CLAUDE_PLUGIN_ROOT}/scripts/claude-seo. These scripts are used to perform web scraping and SEO analysis on user-provided URLs.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external URLs, creating a surface for potential indirect prompt injection attacks where malicious web content could attempt to influence the agent's behavior during comparison or reporting.
  • Ingestion points: HTML elements (Title, Meta descriptions, H1-H3 headings), Open Graph tags, and JSON-LD schema content fetched from external websites.
  • Boundary markers: No explicit delimiters or instruction-ignore warnings are specified for the data processing steps in the skill instructions.
  • Capability inventory: The skill has the ability to execute local scripts, write to a local SQLite database (~/.cache/claude-seo/drift/baselines.db), and generate HTML reports.
  • Sanitization: The instructions explicitly mandate the use of google_auth.validate_url() to prevent SSRF attacks and the use of parameterized placeholders for all SQLite queries to prevent SQL injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 24, 2026, 02:26 AM
Security Audit — agent-trust-hub — seo-drift