* Console to return last processed (#4277)
* Preserve input generator type
* Add tests
* format
---------
Co-authored-by: Jack Gerrits <jack@jackgerrits.com>
Co-authored-by: Eric Zhu <ekzhu@users.noreply.github.com>
* add subscription response
* fix send subscription response
* add register agent type response
* adding a test
* working on shaping up a test
* appsettins update for backend
* another appsettings
* fixup aspire hosting
* enable AGENT_HOST var from aspire
* add SendMessageAsync
* remove broken test
* test compiles and runs but is not (yet) correct
* subscriptions grain wireup.
* temp assert true.
* remove DI for SubscriptionGrain
* add xlang python code
* add subscription response
* rebond
* Update to .NET 9.0
* Fix Backend project SDK
* Package updates
* get RegisterAgentTypeRequest working
* fix exceptions
* add error handling for requests
* whoops
* send cloud event message type
* processing cloudevents
* trying tosend proto data - doesn't work
* trying to pack proto_data
* fix (#4238)
* pack the Message from agents_events
* format - not sure why these?
* format
* cleanup, error handling, xlang sample publishes messages that can be heard by .NET and vice versa
* format
* sdk version
* sdk vers
* net8
* back to net8
* remove netstandard2
* fix used
* remove unused
* more cleanup
* remove unneeded package
* I'm terrible at writing tests
* deserialize the cloud events and sent them as events
* comment
* cleanup
* await
* Delete dotnet/samples/Hello/Backend/Backend.csproj
unneeded change
* whoops
* merge main python back into here
* revert back to local
* revert some of the helloAgents changes.
* [.NET] Add happy path test for in-memory agent && Simplify HelloAgent example && some clean-up in extension APIs (#4227)
* add happy path test
* remove unnecessary namespace
* fix build error
* Update AgentBaseTests.cs
* revert changes
---------
* fix busted merge from main
* addressing review comments
* make internal
* case sensitive rename step 1
* case sensitive rename step 2
* remove!
---------
Co-authored-by: Peter Chang <petchang@microsoft.com>
Co-authored-by: Reuben Bond <reuben.bond@gmail.com>
Co-authored-by: Eric Zhu <ekzhu@users.noreply.github.com>
Co-authored-by: Xiaoyun Zhang <bigmiao.zhang@gmail.com>
* task can be ChatMessage; add group chat state validation step, and address swarm state valiation; change handling of start and reset to RPC to capture any error.
* Update tutorial note
* add ui streaming to distributed group chat example
Signed-off-by: Mohammad Mazraeh <mazraeh.mohammad@gmail.com>
* fix pyright error after updating dependencies
Signed-off-by: Mohammad Mazraeh <mazraeh.mohammad@gmail.com>
---------
Signed-off-by: Mohammad Mazraeh <mazraeh.mohammad@gmail.com>
* add react-flow
* update tutorial, some ui improvement for agent thread layout
* v1 for interactive viz for agent state
* add Ui visualization of message based transitions
* minor updates
* fix node edges, support visualization of self loops
* improve edges and layout, add support for selector group chat prompt
* minor ui tweaks
* ui and layout updates
* ugrade dependencies to fix dependabot scan errors
* persist sidebar, enable contentbar title mechanism #4191
* add support for user proxy agent, support human in put mode. #4011
* add UI support for human input mode via a userproxy agent #4011
* version update
* fix db initialization bug
* support for human input mode in UI, fix backend api route minor bugs
* update pyproject toml and uv lock
* readme update, support full screen mode for agent visualiation
* update uv.lock
When the tool expects complex inputs that require nested models, pydantic will generate a JSON Schema that will include references to other schemas using the `$ref` keyword (cf https://json-schema.org/understanding-json-schema/structuring#dollarref).
But the `$ref` keyword is not supported by OpenAI.
This fix:
1. Resolves the references and inlines the corresponding schemas.
2. Removes the then useless root `$defs` field of the JSON schema to save tokens.
Co-authored-by: Jean-Marc Le Roux <contact@jmlx.io>
Co-authored-by: Jack Gerrits <jackgerrits@users.noreply.github.com>
* first commit
* use_ocr flag
* initial refactor
* test
* small fixes
* adding animations, in progress
* red boundary animations
* add hover tool
* fix timeout time - reduces a lot of errors
* put prompts in separate file
* prompt organize
* add cursor animation
* format and checks pass
* added check for reset
* add react-flow
* update tutorial, some ui improvement for agent thread layout
* v1 for interactive viz for agent state
* add Ui visualization of message based transitions
* minor updates
* fix node edges, support visualization of self loops
* improve edges and layout, add support for selector group chat prompt
* minor ui tweaks
closes#3950closes#3702
What this is doing:
I am refactoring the services on the .NET runtime and attempting to clarify the naming and organization.
I added this doc to help capture the naming and concepts.
AgentRuntime / Worker should work similar to the python version and enables running the whole agent system in one process. For remote the system uses the versions of the services in the grpc folder.
lots of other bug fixes/threading cleanup - passing cancellation token throughout
Services update clarifies the naming and roles:
Worker: Hosts the Agents and is a client to the Gateway
Gateway:
-- RPC gateway for the other services APIs
-- Provides an RPC bridge between the workers and the Event Bus
Registry: keeps track of the agents in the system and which events they can handle
AgentState: persistent state for agents
