gem-hunter

This project creates a small dockerized REST API around signal-cli.

At the moment, the following functionality is exposed via REST:

• Register a number
• Verify the number using the code received via SMS
• Send message (+ attachments) to multiple recipients (or a group)
• Receive messages
• Link devices
• Create/List/Remove groups
• List/Serve/Delete attachments
• Update profile

and many more

1. Create a directory for the configuration This allows you to update signal-cli-rest-api by just deleting and recreating the container without the need to re-register your signal number

$ mkdir -p $HOME/.local/share/signal-api

2. Start a container

$ sudo docker run -d --name signal-api --restart=always -p 8080:8080 \
      -v $HOME/.local/share/signal-api:/home/.local/share/signal-cli \
      -e 'MODE=native' bbernhard/signal-cli-rest-api

3. Register or Link your Signal Number

In this case we'll register our container as secondary device, assuming that you already have your primary number running / assigned to your mobile.

Therefore open http://localhost:8080/v1/qrcodelink?device_name=signal-api in your browser, open Signal on your mobile phone, go to Settings > Linked devices and scan the QR code using the + button.

4. Test your new REST API

Call the REST API endpoint and send a test message: Replace +4412345 with your signal number in international number format, and +44987654 with the recipients number.

$ curl -X POST -H "Content-Type: application/json" 'http://localhost:8080/v2/send' \
     -d '{"message": "Test via Signal API!", "number": "+4412345", "recipients": [ "+44987654" ]}'

You should now have send a message to +44987654.

The signal-cli-rest-api supports three different modes of execution, which can be controlled by setting the MODE environment variable.

• normal Mode: (Default) The signal-cli executable is invoked for every REST API request. Being a Java application, each REST call requires a new startup of the JVM (Java Virtual Machine), increasing the latency and hence leading to the slowest mode of operation.
• native Mode: A precompiled binary signal-cli-native (using GraalVM) is used for every REST API request. This results in a much lower latency & memory usage on each call. On the armv7 platform this mode is not available and falls back to normal. The native mode may also be less stable, due to the experimental state of GraalVM compiler.
• json-rpc Mode: A single, JVM-based signal-cli instance is spawned as daemon process. This mode is usually the fastest, but requires more memory as the JVM keeps running.
• json-rpc-native Mode: Uses the signal-cli-native binary and starts it in daemon mode (this mode basically combines the advantages of the native mode and the json-rpc mode).

Example of running signal-cli-rest in native mode

$ sudo docker run -d --name signal-api --restart=always -p 9922:8080 \
              -v /home/user/signal-api:/home/.local/share/signal-cli \
              -e 'MODE=native' bbernhard/signal-cli-rest-api

This launches an instance of the REST service accessible under http://localhost:9922/v2/send. To preserve the Signal number registration, i.e. for updates, the storage location for the signal-cli configuration is mapped as Docker Volume into a local /home/user/signal-api directory.

⚠️ This setting is only needed in normal/native mode!

signal-cli, which this REST API wrapper is based on, recommends to call receive on a regular basis. So, if you are not already calling the receive endpoint regularly, it is recommended to set the AUTO_RECEIVE_SCHEDULE parameter in the docker-compose.yml file. The AUTO_RECEIVE_SCHEDULE accepts cron schedule expressions and automatically calls the receive endpoint at the given time. e.g: 0 22 * * * calls receive daily at 10pm. If you are not familiar with cron schedule expressions, you can use this website.

WARNING Calling receive will fetch all the messages for the registered Signal number from the Signal Server! So, if you are using the REST API for receiving messages, it's not a good idea to use the AUTO_RECEIVE_SCHEDULE parameter, as you might lose some messages that way.

Sample docker-compose.ymlfile:

version: "3"
services:
  signal-cli-rest-api:
    image: bbernhard/signal-cli-rest-api:latest
    environment:
      - MODE=normal #supported modes: json-rpc, native, normal
      #- AUTO_RECEIVE_SCHEDULE=0 22 * * * #enable this parameter on demand (see description below)
    ports:
      - "8080:8080" #map docker port 8080 to host port 8080.
    volumes:
      - "./signal-cli-config:/home/.local/share/signal-cli" #map "signal-cli-config" folder on host system into docker container. the folder contains the password and cryptographic keys when a new number is registered

The Swagger API documentation can be found here. If you prefer a simple text file based API documentation have a look here.

• Running Signal Messenger REST API in Azure Web App for Containers by @stefanstranger
• Sending Signal Messages by @asad-awadia

In case you need more functionality, please file a ticket or create a PR.

The plugin mechanism allows to register custom endpoints (with different payloads) without forking the project. Have a look here for details.

There are a bunch of environmental variables that can be set inside the docker container in order to change some technical details. This settings are meant for developers and advanced users. Usually you do not need to change anything here - the default values are perfectly fine!

• SIGNAL_CLI_CONFIG_DIR: Specifies the path to the signal-cli config directory inside the docker container. Defaults to /home/.local/share/signal-cli/

• SIGNAL_CLI_UID: Specifies the uid of the signal-api user inside the docker container. Defaults to 1000

• SIGNAL_CLI_GID: Specifies the gid of the signal-api group inside the docker container. Defaults to 1000

• SIGNAL_CLI_CHOWN_ON_STARTUP: If set to false will skip the sometimes time consuming chown on startup. Defaults to true

• SWAGGER_HOST: The host that's used in the Swagger UI for the interactive examples (and useful when this runs behind a reverse proxy). Defaults to SWAGGER_IP:PORT.

• SWAGGER_IP: The IP that's used in the Swagger UI for the interactive examples. Defaults to the container ip.

• SWAGGER_USE_HTTPS_AS_PREFERRED_SCHEME: Use the HTTPS Scheme as preferred scheme in the Swagger UI.

• PORT: Defaults to port 8080 unless this env var is set to tell it otherwise.

• DEFAULT_SIGNAL_TEXT_MODE: Allows to set the default text mode that should be used when sending a message (supported values: normal, styled). The setting is only used in case the text_mode is not explicitly set in the payload of the send method.

• LOG_LEVEL: Allows to set the log level. Supported values: debug, info, warn, error. If nothing is specified, it defaults to info.

• JSON_RPC_IGNORE_ATTACHMENTS: When set to true, attachments are not automatically downloaded in json-rpc mode (default: false)

• JSON_RPC_IGNORE_STORIES: When set to true, stories are not automatically downloaded in json-rpc mode (default: false)

• JSON_RPC_IGNORE_AVATARS: When set to true, avatars are not automatically downloaded in json-rpc mode (default: false)

• JSON_RPC_IGNORE_STICKERS: When set to true, sticker packs are not automatically downloaded in json-rpc mode (default: false)

• JSON_RPC_TRUST_NEW_IDENTITIES: Choose how to trust new identities in json-rpc mode. Supported values: on-first-use, always, never. (default: on-first-use)

• JSON_RPC_RECEIVE_MODE: Controls when signal-cli pulls inbound messages from the Signal servers in json-rpc mode. Supported values: on-start (default), manual. In the default on-start mode, signal-cli auto-receives messages and pushes them as JSON-RPC notifications regardless of whether any websocket subscriber is currently attached to /v1/receive/{number} — messages arriving while no subscriber is attached (e.g. during a brief subscriber redeploy) are silently dropped. In manual mode signal-cli only fetches messages while at least one websocket subscriber is attached; while no subscriber is attached, messages remain queued server-side under Signal's normal retention rules and are delivered on the next subscriber re-attach. Recommended for any deployment where the websocket consumer is restarted or scaled.