redirect-manager

Pass

Audited by Gen Agent Trust Hub on Apr 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local script seo-executor.py for logging and performs modifications to the WordPress .htaccess file for redirect deployment. These are high-impact operations involving local script execution and server configuration changes.
  • [DATA_EXFILTRATION]: The skill accesses local configuration files within ~/.claude-marketing/ to retrieve brand context and compliance rules. It also communicates with external CMS APIs (WordPress/Webflow) and Google Search Console to read and write redirect metadata.
  • [PROMPT_INJECTION]: The skill processes untrusted input through its bulk-import feature, which accepts redirect maps from CSV files or Google Sheets. This represents an indirect prompt injection surface.
  • Ingestion points: User-provided CSV or Google Sheet files used in Step 2.
  • Boundary markers: The skill does not define specific delimiters for imported data but requires an 'approval gate' in Step 5 before execution.
  • Capability inventory: File writing to .htaccess, local script execution of seo-executor.py, and network requests to CMS APIs as described in Steps 6 and 8.
  • Sanitization: The process includes target URL status checks (200 OK) and duplicate rule detection as outlined in Step 2.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 1, 2026, 01:19 AM