NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
NuGet, #2895 most downloaded on NuGet
Classes for loading and saving bot configuration files
Latest stable
4.23.1
Last release 10 months ago
05 Jan 2026
Too new to tell
only 2 release windows
Nearly every release is documented
notes for 1 of 1 stable releases
93 versions withdrawn
withdrawn after publishing
127 years old
94 releases · first in 1900
Drag across the ruler to compare, or tap a release to open it.
Nothing published for this version
Nothing published for this version
Your coding agent can read these notes before it upgrades. Set up the MCP server →
Update ESRP authenticode JSON per SHA1 Digest Deprecation requirement. [[PR 3824]]
This release introduces a global flag called "Templates.EnableFromFile" that indicates whether the Adaptive Expression fromFile function is allowed in LG templates. If an application had previously made use of this function, it is now required to add the line "Templates.EnableFromFile = true;" to the Startup.cs code.
Welcome to the May 2020 release of the Bot Framework SDK. There are a number of updates in this version that we hope you will like; some of the key highlights include:
Insiders: Want to try new features as soon as possible? You can download the nightly Insiders build [C#] [JS] [Python] [CLI] and try the latest updates as soon as they are available. And for the latest Bot Framework news, updates, and content, follow us on Twitter @msbotframework!
Skills have been updated to work with adaptive dialogs, and both adaptive and traditional dialogs will now accept all types of activities.
The skill manifest schema has been updated to version 2.1. Improvements in this version include the ability to declare & share your language models, and define any type of activity that your skill can receive.
This release also includes authentication improvements with skills, including using SSO with dialogs, and OAuth without needing a magic code in WebChat and DirectLine.
We continue to focus on making sure all the Teams-specific APIs are fully supported across our SDKs. This release brings full support for Microsoft Teams APIs in the preview Java SDK, including samples.
The OnTeamsMemberAdded event in the activity handler has been updated to use the get single member endpoint under the covers, which should significantly reduce latency and reliability of this event in large teams.
The TeamsChannelAccount object has been updated to include userRole (one of owner, member, or guest) and tenantId (for the user's tenantId).
Bots now capture Page View events, native to Application Insights, whenever a dialog is started. This allows you to use the User Flows dashboard in Application Insights to see how users move through your bot, between dialogs and where they drop out.
Support was added for a new invoke named healthCheck that allows a sender to verify if a bot is responding to requests, and if trust can be established between the sender and the bot. The bot also has the option of overriding the response to add additional health information in the response.
We’re also excited to make Adaptive Dialogs generally available in C# and as a preview release in JavaScript!
Adaptive Dialogs, which underpin the dialog design and management authoring features found in Bot Framework Composer, enable developers to dynamically update conversation flow based on context and events. This is especially useful when dealing with more sophisticated conversation requirements, such as context switches and interruptions. Bot Framework Skills can now also leverage Adaptive Dialogs.
Adaptive Dialogs also now support Telemetry. Data from Adaptive Dialogs, including triggers, actions and recognizers now flow into your Azure Application Insights instance.
CLI tools for Adaptive Dialogs, Language Generation, QnaMaker and Luis Cross-train - new ability to merge and validate adaptive schema assets, augment qna and lu files, create/ update/ replace/ train/ publish LUIS and or QnA maker application and Language Generation templates manipulation.
New CLI Tools were added for management of Adaptive Dialogs.
LG is Generally Available (GA) on both the C# and JS Platforms.
Language Generation (LG) enables you to define multiple variations of a phrase, execute simple expressions based on context, and refer to conversational memory. At the core of language generation lies template expansion and entity substitution. You can provide one-off variation for expansion as well as conditionally expanding a template. The output from language generation can be a simple text string or multi-line response or a complex object payload that a layer above language generation will use to construct a complete activity. The Bot Framework Composer natively supports language generation to produce output activities using the LG templating system.
You can use Language Generation to:
Language Generation is achieved through:
Adaptive Expressions are Generally Available (GA) on both the C# and JS Platforms.
Bots use expressions to evaluate the outcome of a condition based on runtime information available in memory to the dialog or the Language Generation system. These evaluations determine how your bot reacts to user input and other factors that impact bot functionality.
Adaptive expressions were created to address this core need as well as provide an adaptive expression language that can used with the Bot Framework SDK and other conversational AI components, like Bot Framework Composer, Language Generation, Adaptive dialogs, and Adaptive Cards.
An adaptive expression can contain one or more explicit values, pre-built functions or [custom functions. Consumers of adaptive expressions also have the capability to inject additional supported functions. For example, all Language Generation templates are available as functions as well as additional functions that are only available within that component's use of adaptive expressions.
We added support for single sign-on while using Expect Replies. This applies to SSO performed between a pair of bots: host and a skill.
For Bot Identification we've added the ability to specify sendx5c parameter for certificate authentication. This feature was requested by customers and allows for more flexibility when using cert auth.
Additional Sovereign Clouds are supported.
The Bot Framework has a rich collection of conversational building blocks, but creating a bot that feels natural to converse with requires understanding and coordinating across language understanding, language generation and dialog management. To simplify this process and capture best practices, we've created the bf-generate plugin for the BotFramework CLI tool. The generated dialogs make use of event-driven adaptive dialogs with a rich and evolving set of capabilities including:
Adaptive tools is a brand new Visual studio code extension you can use to create/ validate .lu and .lg documents as well as debug declaratively defined adaptive dialogs. This extension provides rich authoring & editing capabilities for .lu and .lg file formats including syntax highlighting, auto-suggest and auto-complete.
We anticipate adding an early preview to the VS Marketplace shortly after this release.
During this release, the Bot Builder Community has further raised the bar by adding more features, more adapters, and fixing more bugs.
A revised C# Alexa Adapter and Google Home Adapter Re-built from the ground up, starting with Alexa, to allow the adapters to be consumed by Azure Bot Service and made available as channels. Improvements include better native activity type mapping, improved markdown rendering and support for more complex scenarios (such as merging multiple outgoing activities).
A new C# Zoom Adapter. Currently supports Zoom 1:1 and channel chat capabilities, being converted to native BF activity types. Also supports the subscribing to any event a Zoom as supports (translated into Event activities), with full support for Zoom interactive messages and rich message templates.
A RingCentral Adapter. The RingCentral Engage adapter allows you to add an additional endpoint to your bot for RingCentral Engage Digital Platform integration. The RingCentral endpoint can be used in conjunction with other channels meaning, for example, you can have a bot exposed on out of the box channels such as Facebook and Teams, but also integrated as an RingCentral Engage Digital Source SDK into RingCentral.
Nothing published for this version
Nothing published for this version
🪓 BREAKING CHANGES - There is a number of breaking changes. Please see the details here: Adaptive - LG/ expressions
rc0 - this is our first release candidate for Adaptive Dialogs, Adaptive Expressions, & Language generation. We encourage folks to give this a try in development and non-critical scenarios.Expanded support for a number of scenarios in Skills, including
Added support for a new get single member endpoint, and expanded support for the get paged members endpoint.
Nothing published for this version
Nothing published for this version
Nothing published for this version
Welcome to the 4.7 release of the Bot Framework SDK for .NET!
Welcome to the 4.7 release of the Bot Framework SDK for .NET!
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Welcome to the 4.6 release of the Bot Framework SDK for .NET!
Welcome to the 4.6 release of the Bot Framework SDK for .NET!
The Bot Framework SDK v4.6 release fully integrates support for building Teams bots allowing users to use them in channel or group chat conversations. By adding a bot to a team or chat, all users of the conversation can take advantage of the bot functionality right in the conversation.
Create a reusable conversational skill to add functionality to a bot. Leverage pre-built skills, such as Calendar, Email, Task, Point of Interest, Automotive, Weather and News skills. Skills include language models, dialogs, QnA, and integration code delivered to customize and extend as required.
Adaptive Dialogs enable developers to build conversations that can be dynamically changed as the conversation progresses. It allows developers to dynamically update conversation flow based on context and events. This is especially handy when dealing with conversation context switches and interruptions in the middle of a conversation.
Language Generation enables developers to separate logic used to generate bot's respones including ability to define multiple variations on a phrase, execute simple expressions based on context, refer to conversational memory.
Common Expression Language allows you to evaluate the outcome of a condition-based logic at runtime information. We provide a common language that can be used across the Bot Framework SDK and conversational AI components, such as Adaptive Dialogs and Language Generation.
Changes in recognized utterances resulting from consuming different versions of Recognizers-Text through BotBuilder-DotNet. See the changes here.
The issues list can be viewed here.
Nothing published for this version
Support for certificate authentication for AAD apps
This release adds:
C# SDK 4.5.3 release enables cert-based authentication of AAD apps. Currently we only support app id + password, but AAD also supports multiple client credential flows, such as using certificates.
This is the small first step towards using AppCredentials for AAD verifications, which currently supports Certs or Password, but as-is can simply be extended to support any of the 7 ADAL client flows.
The default paths still use ICredentialProvider and we are enabling AppCredentials for new scenarios. A second iteration will make AppCredentials be the default, but we want to stage that transition to minimize risk.
Nothing published for this version
Nothing published for this version
Changed build badge to daily from deprecated CI-CD [[PR 1901]]
Welcome to the 4.5 release of the Bot Framework SDK for .NET!
Today, we are happy to announce the Bot Framework Emulator Channel Testing is generally available. This feature enables developers to debug and test your Bot Framework SDK v4 bots on channels like Microsoft Teams, Slack, Cortana, Facebook Messenger, Skype, etc. As you have the conversation, messages will be mirrored to the Bot Framework Emulator where you can inspect the message data that the bot received. Additionally, a snapshot of the bot state for any given turn between the channel and the bot is rendered as well. You can inspect this data by clicking on the "Bot State" element in the conversation mirror.
We also added capabilities for Unit Testing your bots. The Microsoft.Bot.Builder.Testing package simplifies the process of unit testing dialogs in your bot.
As with any release, we fixed a number of bugs, continue to improve LUIS and QnA integration and further clean our engineering practices. There were additional updates across other areas like Language, Prompts, Dialogs, State and Storage.
Review all changes that went into 4.5 in the detailed Change Log
NextAsync. [PR 2037]LuisRecognizer to handle textless messages [PR 1566]Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Welcome to the 4.4 release of the Bot Framework SDK for .NET!
Welcome to the 4.4 release of the Bot Framework SDK for .NET!
In this release, the team focused on improving and simplifying message and activities handling. The Bot Framework Activity Schema is the underlying schema used to define the interaction model for bots.
As with any release, we fixed a number of bugs, continue to improve LUIS and QnA integration and further clean our engineering practices. There were additional updates across other areas like Language, Prompts, Dialogs, State and Storage.
Review all changes that went into 4.4 in the detailed Change Log
LuisRecognizer to handle textless messages [PR 1566]Nothing published for this version
Nothing published for this version
In this release, the team focused on critical fixes to our token refresh code for OAuth credential flows with Microsoft Application Ids.
In this release, the team focused on critical fixes to our token refresh code for OAuth credential flows with Microsoft Application Ids.
Revamping the token acquisition code that obtains a token given the bot App id and password. This is an effective execution of the OAuth client credentials flow, which is supported by the ADAL dotnet library.
Some implementation details and considerations below to complement the source changes:
Token cache: We use ADAL built-in token cache. Given that Adal handles much of the token caching nowadays, we consider it better to leverage ADAL library implementation, leaving less responsibilities in our code. However, after working with the ADAL team we are still responsible for maintaining low concurrency (done through semaphores in this case) and retrying, using the retry after headers received from AAD when we get HttpStatusCode 429.
Testing: We ran long haul load test, in addition to mini-load tests and unit tests. We were able to reproduce token renewal under extremely high load, and even though there were failures, these were retried and ZERO threads ended up without a token. The longest delay observed in our load test was 600 ms, which happened for a total of 40 requests, then going back to 1 millisecond when reading from cache.
Verifying assumptions Even though the ADAL team recommended limiting concurrency, we removed the semaphore and re-ran the load tests. These results were extremely bad, with latencies up to 5 seconds and multiple failures, confirming ADAL teams' recommendation.
Next steps: We would like to eventually expose the token cache to users so they can provide durable token caches (this is supported by adal). This scenario is awesome for horizontal scaling scenarios where multiple instances share perhaps a redis cache for example.
Coming soon: MSAL library: The ADAL team is working on the MSAL library, which will take care of the concurrency control as well. Eventually we'll move to that. GA for MSAL is coming soon.
In this release, the team focused on enhancing monitoring, telemetry, and analytics capabilities of the SDK by improving the integration with Azure Ap
In this release, the team focused on enhancing monitoring, telemetry, and analytics capabilities of the SDK by improving the integration with Azure App Insights. We have streamlined the integration and default telemetry emitted from the SDK. This include (Waterfall) dialog instrumentation, docs, and examples for querying data, and a PowerBI dashboard. As with any release, we fixed a number of bugs, continue to improve LUIS and QnA integration and further clean our engineering practices. There were additional updates across the other area like Language, Prompt and Dialogs, and Connectors and Adapters.
Review all changes that went into 4.2 in the detailed Change Log See the list of all issues
The 4.1.5 release incorporates changes to the Bot Framework to support activity PR #1078
The 4.1.5 release incorporates changes to the Bot Framework to support activity PR #1078
Authentication in the US Government data centers has also been added.
Nothing published for this version
This release adds support needed for deploying bots to Azure Government based data centers.
This release adds support needed for deploying bots to Azure Government based data centers.
Nothing published for this version
Nothing published for this version
One column per quarter.