email-triage

Pass

Audited by Gen Agent Trust Hub on Jul 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, obfuscation, or unauthorized access were detected in the skill instructions or scripts. The Python script uses standard libraries (argparse, csv, json, re) to perform local data processing.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes external, untrusted data from an email CSV file.
  • Ingestion points: The scripts/email_classifier.py script reads and displays subjects and snippets from inbox.csv.
  • Boundary markers: No explicit delimiters are used in the workflow to separate email content from agent instructions.
  • Capability inventory: The skill allows file reading and local script execution but lacks network or elevated privilege capabilities.
  • Sanitization: Input data is analyzed via regex but not sanitized to remove potentially malicious instructions before being presented to the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 19, 2026, 08:24 PM
Security Audit — agent-trust-hub — email-triage