# Security policy Zero Breach Toolkit is used by people who may be targeted by a state-grade adversary, so its own integrity matters as much as what it detects. ## Reporting a problem Report security issues in the toolkit privately, not as a public issue. That includes any way to make it leak data, contact the network during a scan, read more than it should, misreport a result, or otherwise put a user at risk. - Preferred: open a **private security advisory** on this repository (the "Security" tab, then "Report a vulnerability"). - If you cannot, message the project on Telegram: https://t.me/+t9gq13mSKcE4NGI0 Include what you found, how to reproduce it and the impact. We will acknowledge it, investigate, and credit you if you wish. ## In scope - The scanner reading, decrypting or transmitting anything it should not. - Any path that makes the scanner touch the network during a scan. - A result that is wrong in a way that misleads a user (a false "clean", a mislabelled severity, a missed gap). - The update path (`--update`) fetching from outside its allow-list, or a tampered or changed cached indicator file being accepted at scan time. - The build producing a file that does not match its source. ## Not a vulnerability - **A clean result is not a guarantee.** The tool says so itself. Spyware of this grade is built to defeat retrospective checks; missing a real infection is a documented limitation, not a bug. - A false alarm on a benign indicator match is a tuning issue. Report it as a normal issue. ## Verifying the tool Do not run a security tool on trust. Each release publishes the SHA-256 of its files, and the single-file build is reproducible: you can rebuild `dist/zero-breach-toolkit.py` from source and confirm it matches. Signed releases also include detached GPG signatures made with ArchiveGenocide's established release key; check its full fingerprint against the official publisher page before trusting a downloaded key. See [docs/VERIFYING.md](docs/VERIFYING.md). ## Indicators The indicator set is curated to avoid false alarms against real people: known compromised victim sites and ambiguous legitimate domains are excluded. Historical domains can change ownership later. If an indicator here looks wrong, or could match legitimate traffic, please report it. A false positive from this tool can send an innocent person to a crisis helpline.