# Private Breach Checker

> Check if a password appears in known breaches without revealing it (k-anonymity). Checks a password against the Have I Been Pwned Pwned Passwords collection using k-anonymity: your browser hashes the password with SHA-1 and sends only the first 5 of 40 hex characters; the match is done locally. The server and API accept only a SHA-1 hash or 5-character prefix, never a password.

Part of usefulweb.site (https://usefulweb.site/). Results are deterministic calculations from the inputs and the listed assumptions; the API returns exactly what the web page shows.

## Use it
- Human interface: https://privatebreachchecker.usefulweb.site/ (interactive tool at https://privatebreachchecker.usefulweb.site/tool)
- JSON API: POST https://privatebreachchecker.usefulweb.site/api/v1/calculate with a JSON object of inputs (schema: https://privatebreachchecker.usefulweb.site/api/v1/tool)
- OpenAPI: https://privatebreachchecker.usefulweb.site/openapi.json
- MCP (Streamable HTTP): https://privatebreachchecker.usefulweb.site/mcp - tool `private_breach_checker`

## Inputs
- hash (text, required): SHA-1 hash or 5-character prefix - Hex only. Never enter a password here. With JavaScript on, use the password box above: it is hashed in your browser.

## Method
- The password is hashed with SHA-1 (uppercase hex). Only the first 5 hex characters (20 bits) are sent to https://api.pwnedpasswords.com/range/<prefix>; the service returns every known hash suffix with that prefix (typically 800–1,000) and how often each appeared in breaches.
- Requests use the Add-Padding header, so the response contains random zero-count entries and its size reveals nothing; padding entries are ignored.
- With JavaScript on, hashing, the range request and the comparison all happen in your browser; our server is not involved. If your browser cannot reach the service directly, the page asks our API for the same 5-character range and still compares locally.
- The server path (no JavaScript, API, MCP) accepts only a 40-character SHA-1 hash or a 5-character prefix and rejects anything else, so a password can never be submitted by mistake. With a full hash, the server sees the hash (not the password) and sends only the prefix onwards.
- Not found does not prove a password is strong: it only means it is not in this breach corpus.

## Limits
- If a password appears in a breach, stop using it everywhere and change it. Use a password manager and unique passwords.