Skills
skills/mode-io/mode-io-skills/modeio-anonymization/Gen Agent Trust Hub

modeio-anonymization

Pass

Audited by Gen Agent Trust Hub on Mar 2, 2026

Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted user-provided text for anonymization, which represents a surface for indirect prompt injection.
  • Ingestion points: User content is passed to scripts/anonymize.py and scripts/detect_local.py via the --input command-line argument.
  • Boundary markers: The skill does not utilize specific delimiters or isolation instructions to prevent the agent from potentially obeying instructions embedded within the processed text.
  • Capability inventory: The skill includes network connectivity via scripts/anonymize.py (requests.post).
  • Sanitization: The skill focuses on redacting PII but does not implement sanitization to filter or neutralize malicious prompt injection instructions within the input data.
  • [DATA_EXFILTRATION]: The script scripts/anonymize.py performs a network POST request to https://safety-cf.modeio.ai/api/cf/anonymize. This endpoint belongs to the skill's author ('mode-io') and is the designated destination for the skill's core anonymization functionality.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 2, 2026, 08:39 AM