01 / local
The website never receives your messages.
The CLI reads stable local copies and keeps its database, metrics, profiles, and drafting context on your machine.
About
A local-first CLI and Agent Skill for studying private messaging history and drafting messages that sound like you.
View the open-source project ↗01 / local
The CLI reads stable local copies and keeps its database, metrics, profiles, and drafting context on your machine.
02 / bounded
Deterministic measurements and bounded study packets make the basis of a style profile visible instead of hiding it in model weights.
03 / unsent
Message Like Me produces candidate text for review. It has no command for sending, reacting, scheduling, or operating a messaging account.
Supported inputs
Message Like Me reads Apple Messages locally, caller-owned X archive DMs, bounded Beeper exports, and native WhatsApp evidence made through Ghostget. Optional macOS Contacts adds exact local labels. Each source retains its own provenance and observability limits.
Compare the source boundaries →