geo-platform-optimizer
Pass
Audited by Gen Agent Trust Hub on Oct 20, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill includes scripts/fetch_page.py for retrieving web content. This script implements robust security controls to prevent Server-Side Request Forgery (SSRF), including hostname validation and blocking of non-global (private, loopback, or reserved) IP addresses.
- [INDIRECT_PROMPT_INJECTION]: While the skill processes external content from the web and local files, it contains explicit architectural safeguards. Both SKILL.md and references/skill-contract.md instruct the agent to treat such content as untrusted evidence and specifically direct it to ignore any instructions embedded within that data.
- [COMMAND_EXECUTION]: The skill utilizes several Python scripts for local data processing and reporting. These scripts (brand_scanner.py, citability_scorer.py, generate_markdown_report.py, schema_generator.py) are focused on their stated SEO and marketing tasks and use standard libraries for file and input handling.
- [DYNAMIC_EXECUTION]: A dynamic import (import('sys')) is used in scripts/citability_scorer.py to access standard input when no file argument is provided. This is a common and safe programming pattern in this context and does not involve executing untrusted code.
Audit Metadata