maisecrets / Questions
Questions
The questions people ask most about maisecrets.
Does maisecrets send anything to a server?
No. maisecrets has no telemetry and no server of its own. It sends no prompt and no value. There is no account and no licence check. A connection starts only from a command you run, from an MCP server that you added yourself, or from a NER service that you turned on (off by default): your own ssh command in the Claude Code sandbox goes through the sandbox proxy, and with vault references maisecrets asks that MCP server for the values of its tokens; and a NER service that you turned on gets tool results with the found values as filler, and a typed prompt goes only to a service on your own computer. /maisecrets:report only prints a link, and the connection starts when you open it. The privacy page lists each case. Everything the plugin runs is readable source: Python with only the standard library, plus one small JavaScript mod for Claude Code.
Where does a value go when it is detected?
Into a local vault on your machine: the macOS login keychain, the Windows Credential Locker, or an encrypted file on Linux. Every entry has a time to live. When it expires, the value is deleted.
Is it really free?
Yes. The plugin is Apache-2.0, and every protection in it stays free. mcpgate, the self-hosted gateway for teams, is a separate product.
Can we enforce it for the whole company?
Yes. In a claude.ai organisation an admin adds maisecrets from a private marketplace repository and sets it to required, so every member gets it at session start. In Codex an admin ships the hooks as managed hooks through requirements.toml and MDM; policy trusts them, and users cannot turn them off.
Can I use the detection in my own tool?
Yes, from 0.8.0. The detection core builds as a Python wheel from the repository (it is not on PyPI). A Python program uses maisecrets.detection as a library. A program in any other language sends one JSON request to python3 -m maisecrets.detection and reads one JSON answer (contract maisecrets.detect/1). The core reads no home folder, no configuration and no network.
Does it replace a secrets manager?
No. It keeps values out of the model. A process that runs as you can still read the vault, like any keychain. The threat model in the repository says exactly where the line is.