seo-drift
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external CSV data snapshots from sources such as Google Search Console or rank trackers, which provides an attack surface for indirect instructions.
- Ingestion points: Untrusted data enters the agent context through the files
01-baseline.csvand02-current.csv. - Boundary markers: There are no explicit delimiters or instructions provided to the agent to disregard natural language commands that might be embedded within the CSV data rows.
- Capability inventory: The skill has the ability to execute a local Python script (
seo_drift.py) and chain into other SEO-related tools. - Sanitization: The skill implements a 'Quality scorecard' that checks for data consistency (sample size, metric compatibility, and key collisions), but it does not perform sanitization of text fields to prevent prompt injection.
- [COMMAND_EXECUTION]: The skill executes a local Python script,
scripts/seo_drift.py, to perform the core analytical calculations on the provided datasets.
Audit Metadata