redrover-manager

Fail

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: CRITICALCREDENTIALS_UNSAFEDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • Data Exposure & Exfiltration (HIGH): All script files in the skill utilize require('../../../scripts/secrets.js') to retrieve authentication credentials. Accessing sensitive files containing secrets through relative paths outside the skill directory is a high-risk practice.
  • Data Exposure & Exfiltration (LOW): The skill performs network requests to connect.redroverk12.com. This domain is not included in the trusted whitelist of domains.
  • Indirect Prompt Injection (LOW): The skill possesses an indirect prompt injection surface as it ingests untrusted data from the Red Rover API and outputs it to the console without sanitization. 1. Ingestion points: API data processed in scripts/get_absences.js, scripts/get_daily_summary.js, and scripts/get_weekly_summary.js. 2. Boundary markers: Absent. 3. Capability inventory: Bash and Read tools. 4. Sanitization: Absent.
  • Metadata Poisoning (SAFE): Automated scan alerts identifying location.name as a phishing URL were analyzed and determined to be false positives; the string refers to a JSON property in the API response schema rather than a functional malicious URL.
Recommendations
  • Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata
Risk Level
CRITICAL
Analyzed
Feb 17, 2026, 06:41 PM