Works with the AI Agents you already use
Pii Mask Gate
Masks emails, phone numbers, IBANs, card numbers and IPs in text before you share it, then restores them in the reply. Does not detect names.
$3
pii-mask-gate
Example session with this skill installed
Mask the identifiers in this log before I paste it into an online tool: "Error connecting from 192.168.1.10 for user jane.doe@example.com regarding IBAN FR7630006000011234567890189, call +33 6 12 34 56 78."
- Read your context and instructions
- Compiled the pii-mask-gate
Summary
Masked 4 identifiers locally and verified that unmasking restores the original text exactly. Only the masked text may be shared; the mapping file stays on your machine.
| Type | Placeholder |
|---|---|
| IPv4 address | <IP_ADDRESS_1> |
<EMAIL_1> | |
| IBAN | <IBAN_1> |
| Phone | <PHONE_1> |
Error connecting from <IP_ADDRESS_1> for user <EMAIL_1> regarding IBAN <IBAN_1>, call <PHONE_1>.
Next steps
- Paste only the masked text into the external tool.
- Save the tool's answer and run the unmask command with the same mapping file to restore the originals.
- Note: person names are not detected. Review the text before sharing.
Connects securely to your tools. The creator never sees your data.
About this skill
Replaces structured personal identifiers with typed placeholders (<EMAIL_1>, <PHONE_1>, <IBAN_1>...) before text leaves your machine, and puts the originals back into the answer. Standard-library Python only: no network access, no API key.
What it detects: email addresses, international and French/Moroccan phone numbers, payment card numbers (Luhn-checked), IBANs (mod-97-checked), IPv4 addresses and URLs.
What it does NOT detect: person names, places, organisations, postal addresses and free-text identifiers. In the author's benchmark on synthetic data, the regex detector alone left about 72% of PII characters uncovered, mostly names. The output is masked text, not anonymised text.
Safety behaviour: fail-closed (on any error the script exits non-zero and writes nothing), and the round trip is verified before anything is released. The mapping file stays local.
Origin: a dependency-free extract of the open-source project llm-privacy-gateway (MIT) by Haitam Echchakar, which contains the full benchmark and the CRF and Presidio detectors: https://github.com/echchakarhaitam-cyber/llm-privacy-gateway

Detects structured identifiers only (emails, phone numbers, card numbers, IBANs, IPv4 addresses, URLs). Person names, places, organisations and postal addresses are NOT detected. Phone formats: international numbers with a leading + and French/Moroccan national numbers only.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 5 days ago
Requires
Permissions
Runs a local Python script that reads the text file you give it and writes the masked text and a local mapping file. No network access.