Module prelude

Expand description

The kernel prelude — the ~40 types a typical application touches.

Deliberately a kernel, not an everything-glob. Anything not here lives in a focused submodule and is one import away:

Need	Import
Full Live control plane (persistence, repair, steering, transcripts, contracts)	`use gemini_adk_fluent_rs::live::*;`
Text-agent runtime details	`use gemini_adk_fluent_rs::text::*;`
Toolsets, confirmation, frames	`use gemini_adk_fluent_rs::tools::*;`
State prefixes / `SlotEvidence`	`use gemini_adk_fluent_rs::state::*;`
Full flow vocabulary (`CompiledFlow`, `StepAction`, `Violation`, …)	`use gemini_adk_fluent_rs::flow::*;`
`Agent` trait + operator/pattern internals	`use gemini_adk_fluent_rs::agents::*;`
Conversation compiler (`Conversation`, `ConversationSpec`, …)	`use gemini_adk_fluent_rs::conversation::*;`
A2A, motifs, policy, simulation, testing, orchestration, credentials, run_config	the same-named module, e.g. `use gemini_adk_fluent_rs::simulation::*;`
Raw L0 wire types	`use gemini_adk_fluent_rs::wire::*;`

Re-exports§

pub use crate::voice::Talk;
pub use crate::compose::A;
pub use crate::compose::C;
pub use crate::compose::Ctx;
pub use crate::compose::E;
pub use crate::compose::G;
pub use crate::compose::M;
pub use crate::compose::P;
pub use crate::compose::S;
pub use crate::compose::T;
pub use crate::live::Live;
pub use crate::testing::AgentHarness;
pub use crate::testing::ContractViolation;
pub use crate::testing::DataFlowEdge;
pub use crate::testing::LiveViolation;
pub use crate::testing::check_contracts;
pub use crate::testing::check_live;
pub use crate::testing::diagnose;
pub use crate::testing::infer_data_flow;
pub use crate::builder::*;
pub use crate::operators::*;
pub use crate::patterns::*;

Modules§

tool: Tool dispatch — regular, streaming, and input-streaming tools.
tool_media: Tool media returns — the ADK pattern where a function response carries images or other media to the model alongside its JSON payload, so vision tools (screenshots, chart renderers, document croppers) feed the model something it can actually look at.

Structs§

AudioJitterBuffer: Adaptive jitter buffer for audio playback.
AutomaticActivityDetection: Automatic activity detection (VAD) settings.
BargeInConfig: Configuration for barge-in behavior.
BargeInDetector: Barge-in detector — checks whether user speech should interrupt model output.
BatchJob: A batch prediction job resource.
BatchJobDestination: Destination configuration for a batch job.
BatchJobSource: Source configuration for a batch job.
Blob: A blob of inline data (audio, image, etc.) sent to or received from Gemini.
CachedContent: Cached content resource.
CachedContentUsageMetadata: Usage metadata for cached content.
CallbackContext: Rich context for callbacks — provides access to state, artifacts, memory, and event actions for mutation.
ChatSession: A stateful chat session that tracks conversation history.
CitationMetadata: Citation metadata for a response.
CitationSource: A single citation source.
CodeExecutionResult: Result of code execution.
ConfigError: A build-time configuration error: one or more problems found while validating user-supplied configuration (a Flow, a PhaseMachine, a ComputedRegistry, …).
ConnectBuilder: Connects a SessionConfig with options: transport tuning, a custom transport or codec, a wire recorder. connect is the same thing with none of them.
Content: A content message containing a role and a sequence of parts.
ContentEmbedding: Embedding vector.
ContextWindowCompressionConfig: Context window compression configuration for long sessions.
CountTokensResponse: Response from countTokens.
CreateBatchJobConfig: Configuration for creating a batch job.
CreateCachedContentConfig: Configuration for creating cached content.
CreateTuningJobConfig: Configuration for creating a tuning job.
DispatchTextAgent: Fire-and-forget background task launcher with global task budget.
EmbedContentConfig: Configuration for embed requests.
EmbedContentResponse: Response from embedContent.
EventCallbacks: Typed callback registry for Live session events.
ExecutableCode: Executable code returned by the model.
Extract: A declarative extraction record: typed fields filled by recognizers and/or async resolvers.
FallbackTextAgent: Tries each child agent in sequence. Returns the first successful result. If all fail, returns the last error.
FileData: Reference to an uploaded file.
Flow: A governed conversation/tool DAG.
FlowMonitor: Observes the session trace, maintains the Marking, answers tool admissibility, and projects active postures.
FnTextAgent: Zero-cost state transform agent — executes a closure, no LLM call.
FsPersistence: File-system persistence (good for development and single-server deployments).
FunctionCall: A function call request from the model.
FunctionCallingConfig: Configuration for function calling behavior.
FunctionDeclaration: Schema for a single function that the model can call.
FunctionResponse: A function call response sent back to the model.
GeminiLlm: Concrete Gemini LLM implementation using gemini-live Client.
GeminiLlmParams: Parameters for constructing a GeminiLlm.
GenerateContentConfig: Configuration for a generateContent request.
GenerateContentResponse: Top-level response from generateContent.
GenerationConfig: Generation config sent in the setup message.
GoogleAIAuth: Google AI API key authentication.
GoogleAITokenAuth: Google AI OAuth2 access token authentication.
GoogleSearch: Google Search tool configuration (empty — presence enables the feature).
GoogleSearchRetrieval: Google Search retrieval tool configuration.
GroundingMetadata: Grounding metadata for server content with search results.
HttpClient: HTTP client wrapping reqwest with retry and telemetry.
HttpConfig: Configuration for the HTTP client.
InputAudioTranscription: Input audio transcription configuration.
JitterConfig: Configuration for the jitter buffer.
JoinTextAgent: Waits for dispatched background tasks and collects their results.
JsonCodec: Default JSON codec — current behavior extracted from connection.rs.
ListBatchJobsResponse: Response from listBatchJobs.
ListCachedContentsResponse: Response from listCachedContents.
ListModelsResponse: Response from list models.
ListTuningJobsResponse: Response from listTuningJobs.
LiveHandle: Handle for interacting with a running Live session.
LlmExtractor: LLM-backed turn extractor that sends transcript windows to an OOB LLM with a structured extraction prompt.
LlmTextAgent: Core text agent — calls BaseLlm::generate(), dispatches tools, loops until the model produces a final text response.
LoopTextAgent: Runs a text agent repeatedly until max iterations or a state predicate.
MapOverTextAgent: Iterates a single agent over each item in a state list. Reads state[list_key], runs agent per item (setting state[item_key]), collects results into state[output_key].
MemoryPersistence: In-memory persistence (good for tests).
MockTransport: Mock transport for unit testing.
ModalityTokenCount: Token count breakdown by modality (text, audio, image, video).
ModelId: A Gemini model identifier, exactly as the API accepts it.
OutputAudioTranscription: Output audio transcription configuration.
ParallelTextAgent: Runs text agents concurrently. All branches share state. Results are collected and joined with newlines.
PrebuiltVoiceConfig: Prebuilt voice selection.
ProactivityConfig: Proactivity configuration — controls whether the model can initiate responses.
RaceTextAgent: Runs agents concurrently, returns the first to complete. Cancels the rest.
RealtimeInputConfig: Server-side VAD configuration for the setup message.
RepairConfig: Configuration for the conversation repair system.
ResumeInfo: Session resumption information from the server.
RouteRule: A routing rule: predicate over state → target agent.
RouteTextAgent: State-driven deterministic branching — evaluates predicates in order, dispatches to the first matching agent. Falls back to default if none match.
SafetyRating: Per-category safety assessment of generated content.
SafetySetting: Per-category safety configuration for content generation.
SequentialTextAgent: Runs text agents sequentially. Each agent sees state mutations from previous agents. The final agent’s output is the pipeline’s output.
SessionConfig: Complete session configuration — the builder entrypoint.
SessionHandle: The public API surface for a Gemini Live session.
SessionResumptionConfig: Session resumption configuration.
SimpleTool: Simple function tool that wraps an async closure.
SlidingWindow: Sliding window configuration for context compression.
SoftTurnDetector: Detects proactive silence — user stopped speaking but model didn’t respond.
SpeechConfig: Speech configuration for audio output.
SpscConsumer: The reading half of an SpscRing.
SpscProducer: The writing half of an SpscRing.
SpscRing: Wait-free single-producer single-consumer ring buffer for audio samples.
State: A concurrent, type-safe state container that agents read from and write to.
StateKey: A compile-time typed state key that eliminates typo bugs and type mismatches.
SupervisedTuningSpec: Supervised tuning specification.
TapTextAgent: Read-only observation agent. Calls a function with the state but cannot mutate it. Returns empty string. No LLM call.
TaskRegistry: Shared registry for dispatched background tasks.
TemplateInstruction: A full template-engine instruction (feature templates) — minijinja (Jinja2 syntax: conditionals, loops, filters) over the session state, mirroring ADK’s use_jinja2 instructions. The whole state is exposed as state (subscript prefixed keys: {{ state["session:turn_count"] }}), and every key that is a bare identifier is also available at top level ({{ name }}). Falls back to the empty string for missing values under Jinja’s default undefined semantics.
ThinkingConfig: Configuration for model thinking/reasoning (Gemini 2.5+).
TimeoutTextAgent: Wraps an agent with a time limit. Returns AgentError::Timeout if exceeded.
Tool: A tool declaration sent in the setup message. Each Tool object can contain one of: function declarations, urlContext, googleSearch, codeExecution, or googleSearchRetrieval.
ToolCodeExecution: Code execution tool configuration (empty — presence enables the feature).
ToolConfig: Controls how and when the model uses tools.
ToolContext: Extended context for tool execution — adds function call ID and confirmation.
ToolDispatcher: Routes function calls to the right tool implementation.
ToolPolicy: Optional per-tool execution policy.
TranscriptBuffer: Accumulates input/output transcripts and segments them by turn boundaries.
TranscriptTurn: A single completed conversation turn with accumulated transcripts.
TransportConfig: Configuration for the transport layer.
TungsteniteTransport: WebSocket transport using tokio-tungstenite.
TuningHyperParameters: Tuning hyperparameters.
TuningJob: A tuning job resource.
TurnDetectionConfig: Configuration for client-side turn detection.
TurnDetector: Client-side turn detector.
TypedTool: Type-safe function tool with auto-generated JSON Schema.
UpdateCachedContentRequest: Updates to apply to a cached content resource.
UrlContext: URL context tool configuration (empty — presence enables the feature).
UrlContextMetadata: URL context metadata for content sourced from URLs.
UsageMetadata: Usage metadata returned by the server on messages.
VadConfig: VAD configuration parameters.
VertexAIAuth: Vertex AI Bearer token authentication.
VertexConfig: Configuration for connecting through Vertex AI.
VoiceActivityDetector: Voice Activity Detector with adaptive noise floor.
VoiceConfig: Voice configuration within speech config.

Enums§

AccessToken: An OAuth2 access token source.
ActivityHandling: Controls how incoming audio interacts with model output.
AgentError: Errors that can occur during agent execution.
ApiEndpoint: API endpoint selector — Google AI (direct), Google AI with OAuth token, or Vertex AI.
AudioFormat: Audio encoding formats supported by the Gemini Live API.
AuthError: Authentication-specific errors.
BargeInAction: Result of a barge-in check.
BatchJobState: State of a batch job.
BatchesError: Errors from the Batches API.
BufferState: Current state of the jitter buffer.
CachesError: Errors from the Caches API.
CodecError: Error during encoding or decoding.
ContextDelivery: When to deliver model-role context turns to the wire.
EmbedError: Errors from the Embed API.
EndpointEnvError: Error resolving an ApiEndpoint from the environment.
Enforcement: How a FlowMonitor treats off-path activity — enforcement vs observation.
ExtractionTrigger: Controls WHEN an extractor runs.
FinishReason: Why the model stopped generating.
FunctionCallingBehavior: Whether tool calls block model output or run concurrently.
FunctionCallingMode: How the model should decide when to execute tool calls.
FunctionResponseScheduling: Scheduling mode for non-blocking function responses.
GenerateError: Errors specific to the Generate API.
Guard: A boolean predicate over (state, marking) — the only predicate type.
HarmBlockThreshold: Blocking threshold for safety settings.
HarmCategory: Categories of potential harm in model output.
HarmProbability: Probability that content is harmful.
HttpError: Errors from HTTP client operations.
MediaResolution: Media resolution for image/video inputs.
Modality: Output modalities the model can produce.
ModelsError: Errors from the Models API.
Part: A single part of a Content message. Parts are polymorphic — discriminated by field presence, not a type tag.
PersistenceError: Error returned by a SessionPersistence backend.
Role: Role in a conversation.
Sensitivity: Voice activity detection sensitivity level.
ServerMessage: Unified server message enum — parsed from incoming WebSocket text frames.
ServiceEndpoint: Identifies which Gemini API service to connect to.
SessionCommand: Commands sent from application code to the session transport.
SessionError: Errors that can occur during a session.
SessionEvent: Events emitted by the session, consumed by application code.
SessionPhase: The lifecycle phase of a Gemini Live session.
SetupError: Errors during the setup handshake phase.
SteeringMode: How the phase machine steers the model’s behavior.
TokensError: Errors from the Tokens API.
ToolError: Errors that can occur during tool execution.
TungsteniteError: Errors from the TungsteniteTransport.
TuningJobState: State of a tuning job.
TuningsError: Errors from the Tunings API.
TurnCoverage: Controls which input counts toward a user’s conversation turn.
TurnDetectionEvent: Events from the turn detector.
VadEvent: Events emitted by the VAD.
Verdict: The conformance status of a step.
Voice: Available voice presets for Gemini Live audio output.
WebSocketError: WebSocket-level errors with structured detail.

Constants§

LIVE_INPUT_SAMPLE_RATE: The sample rate the Live API takes on the way in.

Traits§

Agent: The fundamental agent trait. Everything that can process a live session implements this — LLM agents, function agents, pipelines, routers.
AuthProvider: Provides authentication credentials and URL construction for Gemini API connections.
BaseLlm: Trait for LLM providers — decouples agents from specific models.
Codec: Encodes client commands into wire bytes and decodes server bytes into messages.
Frame: A typed conversation frame. Implement via #[derive(Frame)].
InstructionProvider: A dynamic instruction source — the ADK “instruction provider” pattern: instead of a string fixed at build time, the instruction is produced from live session state on every model request (persona switching, risk-driven guardrails, multi-tenant instructions without rebuilding the agent). Any Fn(&State) -> String closure is a provider.
SessionPersistence: Trait for persisting session state across process restarts.
SessionReader: Read-side of a session — subscribe to events and observe phase.
SessionWriter: Write-side of a session — send commands without owning the full handle.
TextAgent: A text-based agent that runs via BaseLlm::generate() (request/response).
ToolFunction: A regular tool — called once, returns a result.
ToolProvider: Declares tools for a Gemini session setup message. Implement this trait to provide tools from any source (runtime ToolDispatcher, etc.).
Transport: A bidirectional message transport.
TurnExtractor: Trait for between-turn extraction from transcript windows.

Functions§

bytes_to_i16: Convert raw bytes to i16 PCM samples (zero-copy via bytemuck).
connect: Connect to the Gemini Multimodal Live API with the default transport and return a session handle.
i16_to_bytes: Convert a slice of i16 PCM samples to raw bytes (zero-copy via bytemuck).
into_shared: Wrap raw bytes as a shared Bytes handle for zero-copy fan-out.
recv_event: Receive the next event from a broadcast receiver, handling lag gracefully.

Type Aliases§

AgentResult: Convenience alias for fallible agent-runtime operations.

Attribute Macros§

tool: Turn an async fn into a registrable Gemini tool.

Derive Macros§

Extract: Derive an Extract record builder from a struct’s fields.
Frame: Derive a Frame impl from a struct’s #[slot(..)] fields.

Module prelude

Module prelude Copy item path

Re-exports§

Modules§

Structs§

Enums§

Constants§

Traits§

Functions§

Type Aliases§

Attribute Macros§

Derive Macros§

Module prelude