juma-geo-audit

Pass

Audited by Gen Agent Trust Hub on Apr 15, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: No malicious patterns, prompt injections, or obfuscation techniques were detected in the provided files. The skill is purely informational and instructional.
  • [NO_CODE]: The skill consists entirely of Markdown files (SKILL.md, citability-scorecard.md, geo-implementation-roadmap.md, and query-test-matrix.md). There are no Python scripts, Node.js code, or shell commands included for execution.
  • [PROMPT_INJECTION]: The instructions focus on auditing brand visibility and do not contain any attempt to override model instructions, bypass safety filters, or extract system prompts.
  • [DATA_EXFILTRATION]: No network-related commands (curl, wget, fetch) or sensitive file access patterns were found. The skill describes a manual documentation process.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 15, 2026, 11:18 AM