Run anonymization checks for text and JSON

Execution policy

1. Default: run scripts/anonymize.py.
2. Use scripts/detect_local.py only if the user explicitly asks for offline, no-network, or local-only detection.

Script commands

Primary mode: scripts/anonymize.py

• -i, --input: content to anonymize
• Script calls the Cloudflare anonymization endpoint internally: https://safety-cf.modeio.ai/api/cf/anonymize.

python scripts/anonymize.py --input "Name: Jack, ID number: 110101199001011234"

python scripts/anonymize.py --input "$(cat sensitive_data.json)"

Output

• data.anonymizedContent: anonymized content string
• data.hasPII: whether sensitive data was detected
• Optional fields include mapping and analysis metadata

Offline mode: scripts/detect_local.py

Use only when user explicitly asks for offline or local detection.

• -i, --input: content to scan
• --json: output full detection details instead of masked text only
• No network call is made.

python scripts/detect_local.py --input "Phone 13812345678 Email test@example.com"

python scripts/detect_local.py --input "Name: Alice Wang, phone 415-555-1234" --json

Output

• sanitizedText: masked text
• items: detected entities
• riskScore: 0-100
• riskLevel: low / medium / high

---

When NOT to use

• For command safety or destructive-operation analysis; use modeio-safety instead.
• For pure policy or legal discussion when no text needs anonymization.

Resources

• scripts/anonymize.py: default script, calls https://safety-cf.modeio.ai/api/cf/anonymize
• scripts/detect_local.py: offline regex detection

Migration context

• Script default endpoint is https://safety-cf.modeio.ai/api/cf/anonymize.
• Anonymization skill traffic is aligned with the Cloudflare rollout.