Hubot adapter for botframework
Перейти к файлу
Melanie Mendoza 17b2f2f785 Merged branch with changes for last reviews 2018-08-24 16:11:22 -07:00
src Removed unneeded commented code 2018-08-24 16:10:14 -07:00
test Changes from latest reviews 2018-08-24 16:08:34 -07:00
.gitignore Adding package-lock.json and bumping version 2017-09-13 14:50:26 -07:00
.travis.yml Adding coveralls 2017-09-18 10:49:27 -07:00
LICENSE.md Rename LICENSE -> LICENSE.md 2017-07-03 14:56:31 -07:00
README.md Some changes for comments and whitespace 2018-08-22 14:29:33 -07:00
coffeelint.json Adding initial system for middlewares 2017-05-22 12:44:57 -07:00
package-lock.json Updated dependencies 2018-08-23 18:27:52 -07:00
package.json Updated dependencies 2018-08-23 18:27:52 -07:00

README.md

Botframework adapter for Hubot

npm version Build Status Coverage Status

Installation

Use hubot in Bot Framework Supported Channels

  1. Install hubot. Make sure to npm install --save hubot-botframework to add this module.

  2. Create a Botframework Registration by completing the Bot Registration Page. Store the created app id and app password for use later.

  3. Configure the required environment variables, and run the command ./bin/hubot -a botframework to run the bot from your local computer.

You can then interact with your hubot through any Bot Framework supported channel.

Additional Steps to Use Hubot in Microsoft Teams

  1. Create a Microsoft Teams app package (.zip) to upload in Teams. We recommend using the manifest editor in App Studio for Microsoft Teams. Include the bot's app id and password in the bots section.

  2. In Microsoft Teams, navigate to the Store and select Upload a custom app. Select the zipped Teams App Package, and install the bot for personal and/or team use.

You can then interact with hubot through a personal chat or by @mentioning the name of the uploaded custom app in a Team. In personal chats, the bot's name can be dropped from messages(ping or hubot ping). In Teams, @mention the bot and omit the bot's name from the command (@myhubot ping).

Global Variables

You can configure the Hubot BotFramework adapter through environment variables.

Required (obtained from the BotFramework portal):

  1. BOTBUILDER_APP_ID - This is the Id of your bot.
  2. BOTBUILDER_APP_PASSWORD - This is the secret for your bot.

Optional:

  1. BOTBUILDER_ENDPOINT - Sets a custom HTTP endpoint for your bot to receive messages on (default is /api/messages).

  2. HUBOT_TEAMS_ENABLE_AUTH - When set to true, restricts sending commands to hubot to a specific set of users in Teams. Messages from all non-Teams channels are blocked. Authorization is disabled by default.

  3. HUBOT_TEAMS_INITIAL_ADMINS - Required if HUBOT_TEAMS_ENABLE_AUTH is true. A comma-separated list of user principal names (UPNs). The users on this list will be admins and able to send commands to hubot when the hubot is first run with authorization enabled.

Channel Specific Variables

Microsoft Teams

These variables will only take effect if a user communicates with your hubot through Microsoft Teams.

Optional:

  1. HUBOT_OFFICE365_TENANT_FILTER - Comma seperated list of Office365 tenant Ids that are allowed to communicate with your hubot. By default ALL Office365 tenants can communicate with your hubot if they sideload your application manifest.

Optional Authorization for Microsoft Teams:

NOTE: The UPNs used for authorization are stored in the hubot brain, so brain persistence affects the use of HUBOT_TEAMS_INITIAL_ADMINS as described below.

Authorization restricts the users that can send commands to hubot to a defined set of Microsoft Teams users. Authorization is currently only supported for the Teams channel, so when enabled, messages from all other channels are blocked. To maximize back compatibility, authorization is disabled by default and must be enabled to be used.

Configuring authorization

Authorization is set up using the HUBOT_TEAMS_ENABLE_AUTH and HUBOT_TEAMS_INITIAL_ADMINS environment variables.

  • HUBOT_TEAMS_ENABLE_AUTH controls whether authorization is enabled or not. If the variable is not set, authorization is disabled. To enable authorization, set the environment variable to true.

  • HUBOT_TEAMS_INITIAL_ADMINS is required if authorization is enabled. This variable contains a comma-separated list of UPNs. When the hubot is run with authorization enabled for the first time, the users whose UPNs are listed will be admins and authorized to send commands to hubot. These UPNs are stored in the hubot brain. After running hubot with authorization enabled for the first time:

    • If your hubot brain is persistent, to change the list of authorized users, first delete the stored list of authorized users from your hubot's brain then change HUBOT_TEAMS_INITIAL_ADMINS to the new list. Also consider using the hubot-msteams script package to dynamically control authorizations.

    • If your hubot brain isn't persistent, the HUBOT_TEAMS_INITIAL_ADMINS list will be used to set admins every time hubot is restarted.

Card-based Interactions for Microsoft Teams

Add screenshots (create an images folder to store them in)

Hubot is great, but hubot without needing to type in whole commands and with less typos is even better. Card-based interactions wrap hubot responses into cards and provide buttons on the card containing useful follow-up commands. To run a follow-up command, simply click the button with the command. If user input is needed, another card is shown with fields for input, and the rest of the command is constructed for you.

Currently, card based interactions are supported for the hubot-github package.

Defining new card-based interactions

Adding new card-based interactions has two steps:

  1. Add entries to HubotResponseCards located in src/hubot-response-cards.coffee. Each entry is from a regex to an array of follow up commands.

    • The regex should map to the command that you want to generate a card for with wildcards for the hubot's name and regexes for each user input. See the hubot-github entries for examples.
    • The follow up queries should match the key for the follow up command in HubotQueryParts.
  2. Add entries to HubotQueryParts located in src/hubot-query-parts.coffee. Each entry is from the command to two arrays containing the text and input parts of a command. These arrays are used to construct the query with any user inputs to send to hubot.

    • textParts contains the text surrounding any user inputs, if a command has no user input, it contains one string in textParts. Note that the first entry of textParts starts with 'hubot'
    • inputParts contains representations of each user input in a command, if any. The text is used to prompt the user for input. A special syntax can used for inputs with finite choices to create a dropdown selector. In this case, a / is used followed by the choices separated by the word " or ". See the hubot-github entries for examples.

Once these entries have been added, cards with follow up commands will be generated for the commands added to HubotResponseCards. For menu cards used to initiate card-based interactions for any command in a script library, use the hubot-msteams library.

Contributing

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments