Overview
The awaken crate is the public facade for the Awaken agent framework. It
re-exports runtime contracts, server contracts, runtime APIs, stores, and
extensions so downstream code can start with one dependency.
Module re-exports
Section titled “Module re-exports”| Facade path | Source crate | Contents |
|---|---|---|
awaken::contract | awaken-runtime-contract | Runtime-facing tools, events, messages, suspension, lifecycle, commit coordinator |
awaken::server_contract | awaken-server-contract | Server/store-facing storage queries, scoped stores, staged commits |
awaken::model | awaken-runtime-contract | Phase, EffectSpec, ScheduledActionSpec, JsonValue |
awaken::registry_spec | awaken-runtime-contract | AgentSpec, ModelSpec, ProviderSpec, McpServerSpec, PluginConfigKey |
awaken::state | awaken-runtime-contract + awaken-runtime | StateKey, StateMap, Snapshot, StateStore, MutationBatch |
awaken::agent | awaken-runtime | Agent configuration and state |
awaken::builder | awaken-runtime | AgentRuntimeBuilder, BuildError |
awaken::context | awaken-runtime | PhaseContext |
awaken::engine | awaken-runtime | LLM engine abstraction |
awaken::execution | awaken-runtime | ExecutionEnv |
awaken::extensions | awaken-runtime | Built-in extension infrastructure |
awaken::loop_runner | awaken-runtime | Agent loop runner |
awaken::phase | awaken-runtime | PhaseRuntime, PhaseHook |
awaken::plugins | awaken-runtime | Plugin, PluginDescriptor, PluginRegistrar |
awaken::policies | awaken-runtime | Context window and retry policies |
awaken::registry | awaken-runtime | AgentResolver, ResolvedAgent, ResolvedBackendAgent |
awaken::runtime | awaken-runtime | AgentRuntime |
awaken::stores | awaken-stores | Memory, file, PostgreSQL, and SQLite-backed store implementations |
Feature-gated modules
Section titled “Feature-gated modules”| Facade path | Feature flag | Source crate |
|---|---|---|
awaken::ext_permission | permission | awaken-ext-permission |
awaken::ext_observability | observability | awaken-ext-observability |
awaken::ext_mcp | mcp | awaken-ext-mcp |
awaken::ext_skills | skills | awaken-ext-skills |
awaken::ext_generative_ui | generative-ui | awaken-ext-generative-ui |
awaken::ext_reminder | reminder | awaken-ext-reminder |
awaken::server | server | awaken-server |
Root-level re-exports
Section titled “Root-level re-exports”The following types are re-exported at the crate root for convenience:
From awaken-runtime-contract:
AgentSpec, EffectSpec, FailedScheduledActions, JsonValue, KeyScope,
MergeStrategy, PendingScheduledActions, PersistedState, Phase,
PluginConfigKey, ScheduledActionSpec, Snapshot, StateError, StateKey,
StateKeyOptions, StateMap, TypedEffect, UnknownKeyPolicy
From awaken-runtime:
AgentResolver, AgentRuntime, AgentRuntimeBuilder, BuildError,
CancellationToken, CommitEvent, CommitHook, DEFAULT_MAX_PHASE_ROUNDS,
ExecutionEnv, MutationBatch, PhaseContext, PhaseHook, PhaseRuntime,
Plugin, PluginDescriptor, PluginRegistrar, ResolvedAgent, RunActivation,
RuntimeError, StateCommand, StateStore, ToolGateHook,
TypedEffectHandler, TypedScheduledActionHandler
Feature flags
Section titled “Feature flags”| Flag | Default | Description |
|---|---|---|
permission | yes | Tool-level permission gating (HITL) |
observability | yes | Tracing and metrics integration |
mcp | yes | MCP (Model Context Protocol) tool bridge |
skills | yes | Skills subsystem for reusable agent capabilities |
reminder | yes | Reminder extension for injecting context messages |
server | yes | HTTP server with SSE streaming and protocol adapters |
generative-ui | yes | Generative UI component streaming |
full | yes | Enables all of the above |
Workspace extension crates can exist outside the facade feature set. The current
one is awaken-ext-deferred-tools; add it as a direct dependency when you need
deferred tool loading.