Merge pull request deepgram#520 from deepgram/feat/agent-v1

feat: support for agent v1

Author: naomi-lgbt

.github/CONTRIBUTING.md

@@ -48,3 +48,18 @@ Here are a few types of contributions that we would be interested in hearing abo
 ## Making Code Contributions
 
 for those interested in contributing code to the project, please review the [Code Contribution Guide](https://github.com/deepgram/deepgram-python-sdk/blob/main/.github/CODE_CONTRIBUTIONS_GUIDE.md) for more details.
+
+## Building Locally
+
+Assuming you are using `pipenv`:
+
+```bash
+# Install deps
+pipenv install
+# Build package
+pipenv run python3 -m build
+# Install package from local build
+pipenv install ./dist/deepgram_sdk-0.0.0.tar.gz
+# Try an example!
+DEEPGRAM_API_KEY=<key> pipenv run python3 examples/agent/async_simple/main.py
+```

.gitignore

@@ -21,3 +21,6 @@ dist/
 build/
 poetry.lock
 
+# examples
+chatlog.txt
+output_*.wav

deepgram/__init__.py

@@ -324,7 +324,6 @@
  ConversationTextResponse,
  UserStartedSpeakingResponse,
  AgentThinkingResponse,
- FunctionCalling,
  FunctionCallRequest,
  AgentStartedSpeakingResponse,
  AgentAudioDoneResponse,
@@ -333,27 +332,29 @@
 
 from .client import (
  # top level
- SettingsConfigurationOptions,
- UpdateInstructionsOptions,
+ SettingsOptions,
+ UpdatePromptOptions,
  UpdateSpeakOptions,
  InjectAgentMessageOptions,
  FunctionCallResponse,
  AgentKeepAlive,
  # sub level
  Listen,
+ ListenProvider,
  Speak,
+ SpeakProvider,
  Header,
  Item,
  Properties,
  Parameters,
  Function,
- Provider,
  Think,
+ ThinkProvider,
  Agent,
  Input,
  Output,
  Audio,
- Context,
+ Endpoint,
 )
 
 # utilities

deepgram/client.py

@@ -338,7 +338,6 @@
  ConversationTextResponse,
  UserStartedSpeakingResponse,
  AgentThinkingResponse,
- FunctionCalling,
  FunctionCallRequest,
  AgentStartedSpeakingResponse,
  AgentAudioDoneResponse,
@@ -347,27 +346,29 @@
 
 from .clients import (
  # top level
- SettingsConfigurationOptions,
- UpdateInstructionsOptions,
+ SettingsOptions,
+ UpdatePromptOptions,
  UpdateSpeakOptions,
  InjectAgentMessageOptions,
  FunctionCallResponse,
  AgentKeepAlive,
  # sub level
  Listen,
+ ListenProvider,
  Speak,
+ SpeakProvider,
  Header,
  Item,
  Properties,
  Parameters,
  Function,
- Provider,
  Think,
+ ThinkProvider,
  Agent,
  Input,
  Output,
  Audio,
- Context,
+ Endpoint,
 )
 
 

deepgram/clients/__init__.py

@@ -347,7 +347,6 @@
  ConversationTextResponse,
  UserStartedSpeakingResponse,
  AgentThinkingResponse,
- FunctionCalling,
  FunctionCallRequest,
  AgentStartedSpeakingResponse,
  AgentAudioDoneResponse,
@@ -356,25 +355,27 @@
 
 from .agent import (
  # top level
- SettingsConfigurationOptions,
- UpdateInstructionsOptions,
+ SettingsOptions,
+ UpdatePromptOptions,
  UpdateSpeakOptions,
  InjectAgentMessageOptions,
  FunctionCallResponse,
  AgentKeepAlive,
  # sub level
  Listen,
+ ListenProvider,
  Speak,
+ SpeakProvider,
  Header,
  Item,
  Properties,
  Parameters,
  Function,
- Provider,
  Think,
+ ThinkProvider,
  Agent,
  Input,
  Output,
  Audio,
- Context,
+ Endpoint,
 )

deepgram/clients/agent/__init__.py

@@ -22,7 +22,6 @@
  ConversationTextResponse,
  UserStartedSpeakingResponse,
  AgentThinkingResponse,
- FunctionCalling,
  FunctionCallRequest,
  AgentStartedSpeakingResponse,
  AgentAudioDoneResponse,
@@ -31,25 +30,27 @@
 
 from .client import (
  # top level
- SettingsConfigurationOptions,
- UpdateInstructionsOptions,
+ SettingsOptions,
+ UpdatePromptOptions,
  UpdateSpeakOptions,
  InjectAgentMessageOptions,
  FunctionCallResponse,
  AgentKeepAlive,
  # sub level
  Listen,
+ ListenProvider,
  Speak,
+ SpeakProvider,
  Header,
  Item,
  Properties,
  Parameters,
  Function,
- Provider,
  Think,
+ ThinkProvider,
  Agent,
  Input,
  Output,
  Audio,
- Context,
+ Endpoint,
 )

deepgram/clients/agent/client.py

@@ -21,7 +21,6 @@
  ConversationTextResponse as LatestConversationTextResponse,
  UserStartedSpeakingResponse as LatestUserStartedSpeakingResponse,
  AgentThinkingResponse as LatestAgentThinkingResponse,
- FunctionCalling as LatestFunctionCalling,
  FunctionCallRequest as LatestFunctionCallRequest,
  AgentStartedSpeakingResponse as LatestAgentStartedSpeakingResponse,
  AgentAudioDoneResponse as LatestAgentAudioDoneResponse,
@@ -30,27 +29,29 @@
 
 from .v1 import (
  # top level
- SettingsConfigurationOptions as LatestSettingsConfigurationOptions,
- UpdateInstructionsOptions as LatestUpdateInstructionsOptions,
+ SettingsOptions as LatestSettingsOptions,
+ UpdatePromptOptions as LatestUpdatePromptOptions,
  UpdateSpeakOptions as LatestUpdateSpeakOptions,
  InjectAgentMessageOptions as LatestInjectAgentMessageOptions,
  FunctionCallResponse as LatestFunctionCallResponse,
  AgentKeepAlive as LatestAgentKeepAlive,
  # sub level
  Listen as LatestListen,
+ ListenProvider as LatestListenProvider,
  Speak as LatestSpeak,
+ SpeakProvider as LatestSpeakProvider,
  Header as LatestHeader,
  Item as LatestItem,
  Properties as LatestProperties,
  Parameters as LatestParameters,
  Function as LatestFunction,
- Provider as LatestProvider,
  Think as LatestThink,
+ ThinkProvider as LatestThinkProvider,
  Agent as LatestAgent,
  Input as LatestInput,
  Output as LatestOutput,
  Audio as LatestAudio,
- Context as LatestContext,
+ Endpoint as LatestEndpoint,
 )
 
 
@@ -70,31 +71,32 @@
 ConversationTextResponse = LatestConversationTextResponse
 UserStartedSpeakingResponse = LatestUserStartedSpeakingResponse
 AgentThinkingResponse = LatestAgentThinkingResponse
-FunctionCalling = LatestFunctionCalling
 FunctionCallRequest = LatestFunctionCallRequest
 AgentStartedSpeakingResponse = LatestAgentStartedSpeakingResponse
 AgentAudioDoneResponse = LatestAgentAudioDoneResponse
 InjectionRefusedResponse = LatestInjectionRefusedResponse
 
 
-SettingsConfigurationOptions = LatestSettingsConfigurationOptions
-UpdateInstructionsOptions = LatestUpdateInstructionsOptions
+SettingsOptions = LatestSettingsOptions
+UpdatePromptOptions = LatestUpdatePromptOptions
 UpdateSpeakOptions = LatestUpdateSpeakOptions
 InjectAgentMessageOptions = LatestInjectAgentMessageOptions
 FunctionCallResponse = LatestFunctionCallResponse
 AgentKeepAlive = LatestAgentKeepAlive
 
 Listen = LatestListen
+ListenProvider = LatestListenProvider
 Speak = LatestSpeak
+SpeakProvider = LatestSpeakProvider
 Header = LatestHeader
 Item = LatestItem
 Properties = LatestProperties
 Parameters = LatestParameters
 Function = LatestFunction
-Provider = LatestProvider
 Think = LatestThink
+ThinkProvider = LatestThinkProvider
 Agent = LatestAgent
 Input = LatestInput
 Output = LatestOutput
 Audio = LatestAudio
-Context = LatestContext
+Endpoint = LatestEndpoint

deepgram/clients/agent/enums.py

@@ -21,16 +21,15 @@ class AgentWebSocketEvents(StrEnum):
  ConversationText: str = "ConversationText"
  UserStartedSpeaking: str = "UserStartedSpeaking"
  AgentThinking: str = "AgentThinking"
- FunctionCalling: str = "FunctionCalling"
  FunctionCallRequest: str = "FunctionCallRequest"
  AgentStartedSpeaking: str = "AgentStartedSpeaking"
  AgentAudioDone: str = "AgentAudioDone"
  Error: str = "Error"
  Unhandled: str = "Unhandled"
 
  # client
- SettingsConfiguration: str = "SettingsConfiguration"
- UpdateInstructions: str = "UpdateInstructions"
+ Settings: str = "Settings"
+ UpdatePrompt: str = "UpdatePrompt"
  UpdateSpeak: str = "UpdateSpeak"
  InjectAgentMessage: str = "InjectAgentMessage"
  InjectionRefused: str = "InjectionRefused"

deepgram/clients/agent/v1/__init__.py

@@ -26,7 +26,6 @@
  ConversationTextResponse,
  UserStartedSpeakingResponse,
  AgentThinkingResponse,
- FunctionCalling,
  FunctionCallRequest,
  AgentStartedSpeakingResponse,
  AgentAudioDoneResponse,
@@ -35,25 +34,27 @@
 
 from .websocket import (
  # top level
- SettingsConfigurationOptions,
- UpdateInstructionsOptions,
+ SettingsOptions,
+ UpdatePromptOptions,
  UpdateSpeakOptions,
  InjectAgentMessageOptions,
  FunctionCallResponse,
  AgentKeepAlive,
  # sub level
  Listen,
+ ListenProvider,
  Speak,
+ SpeakProvider,
  Header,
  Item,
  Properties,
  Parameters,
  Function,
- Provider,
  Think,
+ ThinkProvider,
  Agent,
  Input,
  Output,
  Audio,
- Context,
+ Endpoint,
 )

deepgram/clients/agent/v1/websocket/__init__.py

@@ -18,33 +18,34 @@
  ConversationTextResponse,
  UserStartedSpeakingResponse,
  AgentThinkingResponse,
- FunctionCalling,
  FunctionCallRequest,
  AgentStartedSpeakingResponse,
  AgentAudioDoneResponse,
  InjectionRefusedResponse,
 )
 from .options import (
  # top level
- SettingsConfigurationOptions,
- UpdateInstructionsOptions,
+ SettingsOptions,
+ UpdatePromptOptions,
  UpdateSpeakOptions,
  InjectAgentMessageOptions,
  FunctionCallResponse,
  AgentKeepAlive,
  # sub level
  Listen,
+ ListenProvider,
  Speak,
+ SpeakProvider,
  Header,
  Item,
  Properties,
  Parameters,
  Function,
- Provider,
  Think,
+ ThinkProvider,
  Agent,
  Input,
  Output,
  Audio,
- Context,
+ Endpoint,
 )