Draft — last updated June 11, 2026
HumanifyMe is built local-first specifically so that we don't have to ask you to trust us. This page says in plain English what the software does with data. The short version: we collect nothing, because there is nothing for us to collect.
Nothing. HumanifyMe has no accounts, no telemetry, no analytics, and no backend server operated by us. We cannot see your writing samples, your style profile, your drafts, or your rewrites. The software runs entirely on your computer.
~/.humanifyme/data.db.~/.humanifyme/config.json, with file permissions restricted to your user account.Two operations send text over the network, both to the LLM provider you configure (Anthropic, OpenAI, Google Gemini, or a local Ollama server) and to no other destination:
Redaction masks email addresses, phone numbers, street addresses, card numbers, API keys, and tokens before any text is sent, and restores them in your local copy afterward. If you use Ollama, nothing leaves your machine at all.
Your relationship with your LLM provider is governed by their terms. We default to providers' no-training API tiers where they exist and document provider data-use policies in the project repository.
Run humanifyme wipe --confirm (or call the humanify_wipe_all tool from your agent). This deletes your samples, profile, cache, and audit log irrevocably. Because nothing is stored anywhere else, that's genuinely everything.
If we later offer an optional managed service (e.g., cross-device sync), it will be opt-in, end-to-end encrypted with a key derived on your device, and covered by an updated version of this policy before launch. Local-only operation will remain available.