A plugin that logs important information at each callback point.
This plugin helps print all critical events in the console.
A plugin that logs a high volume of requests and responses handled by the agent at each callback point.
This plugin is primarily intended for ADK development and debugging purposes, helping to print all critical events in the console.
CAUTION: The plugin logs raw requests / responses, including user prompts. Be mindful of sensitive data disclosure.
A plugin that logs important information at each callback point.
A plugin that logs a high volume of requests and responses handled by the agent at each callback point.
Parses a JSON string to a map.
Parses a JSON string to a map.
Parses a JSON string to a map.
Returns the underlying function declaration.
Returns the underlying function declaration.
A built-in tool that is automatically invoked by Gemini 2 models to retrieve search results from Google Maps.
This tool operates internally within the model and does not require or perform local code execution.
A built-in tool that is automatically invoked by Gemini 2 models to retrieve search results from Google Maps.
This tool operates internally within the model and does not require or perform local code execution.
The custom metadata of the tool.
The custom metadata of the tool.
The description of the tool.
The description of the tool.
Whether the tool's final result will be delivered out-of-band. When true, the framework marks the call as long-running and uses the tool's return value as the function-response payload (or suppresses the response entirely if the tool returns Unit).
Whether the tool's final result will be delivered out-of-band. When true, the framework marks the call as long-running and uses the tool's return value as the function-response payload (or suppresses the response entirely if the tool returns Unit).
Returns the underlying function declaration.
Returns the underlying function declaration.
Processes the LLM request before it is sent.
Processes the LLM request before it is sent.
Processes the LLM request before it is sent.
Tools can override this to attach instructions, artifacts, or other data to the request. By default, this implementation appends the tool itself to the LlmRequest, making it available for use by the LLM.
Processes the LLM request before it is sent.
Tools can override this to attach instructions, artifacts, or other data to the request. By default, this implementation appends the tool itself to the LlmRequest, making it available for use by the LLM.
Returns the underlying function declaration.
Returns the underlying function declaration.
A built-in tool that is automatically invoked by Gemini 2 and 3 models to retrieve search results from Google Search.
This tool operates internally within the model and does not require or perform local code execution.
A built-in tool that is automatically invoked by Gemini 2 and 3 models to retrieve search results from Google Search.
This tool operates internally within the model and does not require or perform local code execution.
The custom metadata of the tool.
The custom metadata of the tool.
The description of the tool.
The description of the tool.
Whether the tool's final result will be delivered out-of-band. When true, the framework marks the call as long-running and uses the tool's return value as the function-response payload (or suppresses the response entirely if the tool returns Unit).
Whether the tool's final result will be delivered out-of-band. When true, the framework marks the call as long-running and uses the tool's return value as the function-response payload (or suppresses the response entirely if the tool returns Unit).
Returns the underlying function declaration.
Returns the underlying function declaration.
Processes the LLM request before it is sent.
Processes the LLM request before it is sent.
Processes the LLM request before it is sent.
Tools can override this to attach instructions, artifacts, or other data to the request. By default, this implementation appends the tool itself to the LlmRequest, making it available for use by the LLM.
Processes the LLM request before it is sent.
Tools can override this to attach instructions, artifacts, or other data to the request. By default, this implementation appends the tool itself to the LlmRequest, making it available for use by the LLM.
Returns the underlying function declaration.
Returns the underlying function declaration.
A built-in tool using Vertex AI Search.
This tool can be configured with either a dataStoreId (the Vertex AI search data store resource ID) or a searchEngineId (the Vertex AI search engine resource ID).
A built-in tool using Vertex AI Search.
This tool can be configured with either a dataStoreId (the Vertex AI search data store resource ID) or a searchEngineId (the Vertex AI search engine resource ID).
The custom metadata of the tool.
The custom metadata of the tool.
The Vertex AI search data store resource ID in the format of projects/{project}/locations/{location}/collections/{collection}/dataStores/{dataStore}.
The Vertex AI search data store resource ID in the format of projects/{project}/locations/{location}/collections/{collection}/dataStores/{dataStore}.
Specifications that define the specific DataStores to be searched. It should only be set if engine is used.
Specifications that define the specific DataStores to be searched. It should only be set if engine is used.
The description of the tool.
The description of the tool.
Whether the tool's final result will be delivered out-of-band. When true, the framework marks the call as long-running and uses the tool's return value as the function-response payload (or suppresses the response entirely if the tool returns Unit).
Whether the tool's final result will be delivered out-of-band. When true, the framework marks the call as long-running and uses the tool's return value as the function-response payload (or suppresses the response entirely if the tool returns Unit).
The maximum number of results to return.
The maximum number of results to return.
The model name to use, overriding the one in LlmRequest.
The model name to use, overriding the one in LlmRequest.
The Vertex AI search engine resource ID in the format of projects/{project}/locations/{location}/collections/{collection}/engines/{engine}.
The Vertex AI search engine resource ID in the format of projects/{project}/locations/{location}/collections/{collection}/engines/{engine}.
Returns the underlying function declaration.
Returns the underlying function declaration.
Processes the LLM request before it is sent.
Processes the LLM request before it is sent.
Processes the LLM request before it is sent.
Tools can override this to attach instructions, artifacts, or other data to the request. By default, this implementation appends the tool itself to the LlmRequest, making it available for use by the LLM.
Processes the LLM request before it is sent.
Tools can override this to attach instructions, artifacts, or other data to the request. By default, this implementation appends the tool itself to the LlmRequest, making it available for use by the LLM.
A built-in tool that is automatically invoked by Gemini 2 models to retrieve search results from Google Maps.
A built-in tool that is automatically invoked by Gemini 2 models to retrieve search results from Google Maps.
A built-in tool that is automatically invoked by Gemini 2 and 3 models to retrieve search results from Google Search.
A built-in tool that is automatically invoked by Gemini 2 and 3 models to retrieve search results from Google Search.
A built-in tool using Vertex AI Search.
A built-in tool using Vertex AI Search.
Utility class for validating schemas.
Validates a map against a schema.
Result.success if the map matches the schema, Result.failure wrapping an IllegalArgumentException describing the first validation error otherwise.
The map to validate.
The schema to validate against.
The name of the arguments being validated (e.g., "Input" or "Output").
Utility class for validating schemas.
A service that allows the agent to save a generated report directly to the ArtifactService.
Retrieves a previously saved markdown report from the artifact service.
The filename to retrieve. Should end in .md
Saves a generated markdown report to the artifact service.
The desired filename for the report. Should end in .md
The full markdown content of the generated report.
Example agent demonstrating a common use case for the Artifact Service: Generated Reports.
This example showcases:
Defining a custom service using @Tool annotations that invokes the ArtifactService to save its output.
An agent generating a report and using the tool to "persist" the data.
The overarching application using GcsArtifactService as the backend storage provider.
A service that allows the agent to save a generated report directly to the ArtifactService.
Example agent demonstrating a common use case for the Artifact Service: Generated Reports.
Whimsical demo agent Alice in Processorland demonstrating callbacks and observers.
A callback that logs a Cheshire Cat grin before the model runs!
Shrinks the user so they can fit through small doors.
Humorous Alice-in-Wonderland tools exposed to the LLM via @Tool annotations.
Whimsical demo agent Alice in Processorland demonstrating callbacks and observers.
A callback that logs a Cheshire Cat grin before the model runs!
A callback that warns about being late before a tool is called!
Humorous Alice-in-Wonderland tools exposed to the LLM via @Tool annotations.
A fun, nerdy demo for LoopAgent simulating an RPG loot grinder.
A fun, quirky, family-friendly demo for ParallelAgent showcasing simultaneous actions by a team of cats!
A fun, nerdy demo for SequentialAgent simulating a pre-flight sequence.
A fun, nerdy demo for LoopAgent simulating an RPG loot grinder.
A fun, quirky, family-friendly demo for ParallelAgent showcasing simultaneous actions by a team of cats!
A fun, nerdy demo for SequentialAgent simulating a pre-flight sequence.
Use constructor without runner
AdkWebServer(port, sessionService, artifactService, agentLoader, apiServerSpanExporter)Use adkModule without runner
adkModule(sessionService, artifactService, agentLoader, apiServerSpanExporter)