Stable API Exports

275 exports guaranteed stable across minor releases. Source: docs/api/stable-api-snapshot.json. Breaking removals only in major releases.

A

• ActionLanguage
• ActionStep
• AdaptationDecision
• AdaptationProposal
• AgentEvent
• AgentPrincipal
• AgentRunConfig
• AgentSnapshot
• AgentSupervisor
• AgentSupervisorOptions
• AgentTeam
• AgentTeamFactory
• AgentTeamMember
• AgentTeamMemberResult
• AgentTeamOptions
• AgentTeamResult
• AgentTeamScorer
• AgentTeamScorerInput
• AgentTeamSpawnContext
• AnthropicModel
• AnthropicModelId
• AnthropicModelOptions
• AnthropicModels
• ApplyStateDeltaOptions
• ApprovalPolicy
• ApprovalPolicyOptions
• ApprovalRule
• AsToolOptions
• AssemblerConfig
• adaptation
• agents
• agui
• applyApprovalPolicy
• applyHumanResponse
• applyStateDelta
• approvalPolicy
• asTool
• assertPathAllowed
• axis

B

• Bm25Indexer
• Bm25Match
• BuildFrontendToolsOptions
• BuildPassesVerifier
• BuildPassesVerifierOptions
• BuildResult
• BuildResultReader
• BuildStatus
• bm25Tokenize
• budgetGuardPolicy
• buildCapabilityGlobals
• buildFrontendTools
• buildSandboxFetch

C

• CACHE_MIN_TOKENS
• CapabilityManifest
• CheckpointableAgentOptions
• CheckpointableRun
• Checkpointer
• ClassifierGuardrailOptions
• CodeAgent
• CodeAgentOptions
• CodeGuardrailOptions
• ContentBlock
• Criterion
• CriterionVerdict
• callFingerprint
• checkpoint
• classifierGuardrail
• codeGuardrail
• composePolicies
• coreMemoryTools
• costBudget
• createKernel
• createMemoryTool

D

• DEFAULT_CRITERIA_SYNTH_SYSTEM_PROMPT
• DecayOptions
• DecayResult
• DeterministicVerifier
• denyTools

E

• EditToolResultsOptions
• EmbedResult
• Embedder
• EnhancementPolicy
• EventLog
• EventLogOptions
• estimateMessagesTokens
• estimateTokens
• executor
• extractActionsFromResponse

F

• FallbackModel
• FallbackModelOptions
• FinalAnswerStep
• FrontendToolDispatcher
• FrontendToolSpec
• forbiddenPhrases
• formatSseFrame
• frontendTools

G

• GenerateOptions
• GenericOpenAICompatModel
• GenericOpenAICompatModelOptions
• Goal
• GoalAgent
• GoalAgentOptions
• GoalDirectedAgent
• GoalDirectedAgentOptions
• GoalDirectedOutcome
• GoalDirectedRunResult
• GoalOutcome
• GoalRunResult
• GuardrailResult
• guardrails

H

• HandoffAgent
• HandoffOptions
• HandoffResult
• HybridRetriever
• HybridRetrieverOpts
• handoff
• handoffGenerator
• hybridRetriever

I

• ImageBlock
• InMemoryCheckpointer
• InMemoryStructuredKv
• InMemoryVectorStore
• InputGuardrail
• IntentAlignmentGuardrailOptions
• index
• intentAlignmentGuardrail

J

• JsKernel

K

• KernelEngine
• KernelOptions
• KernelPool
• KernelPoolOptions
• KernelPoolValidator
• KernelPoolValidatorOptions
• KernelResult
• KernelValidationResult
• KvBackend
• KvBackendVectorStore
• KvCheckpointer

L

• L3
• LLMJudgeVerifier
• LLMJudgeVerifierOptions
• LLM_JUDGE_SYSTEM_PROMPT
• LazyObservationHandle
• LoggedEvent
• llamaGuardAdapter
• longestAnswerScorer

M

• MapKvBackend
• McpAuthError
• McpAuthOptions
• McpGetPromptResult
• McpIntegrityOptions
• McpPromptMessage
• McpPromptSchema
• McpResource
• McpResourceContent
• McpToolCollection
• McpToolSchema
• MemoryBlock
• MemoryBlockSet
• MemoryEntry
• MemoryNamespace
• MemoryToolOptions
• MessageAssembler
• Model
• ModelCapabilities
• ModelMessage
• makeRetrievalTool
• matchGlob
• maxInputLength
• memory
• models

N

• negotiation
• noProgress
• noProgressPolicy

O

• Observation
• ObservationPriority
• ObservationalMemory
• ObservationalMemoryOptions
• OpenAICompatModel
• OpenAICompatModelOptions
• OpenAIModel
• OpenAIModelId
• OpenAIModelOptions
• OpenAIModels
• OutputGuardrail

P

• PAIRWISE_JUDGE_SYSTEM_PROMPT
• PairwiseVerdict
• ParallelToolUseCall
• ParallelToolUseStep
• ParsedAction
• ParsedActionType
• PlanningStep
• PolicyPresets
• ProgrammaticOrchestrator
• ProgrammaticResult
• parseCriteriaReply
• parseStopPolicies
• parseStopPolicy
• policies

Q

• QueryFilter

R

• RedisClientLike
• RedisClientOptions
• RedisKvBackend
• RedisRestKvBackend
• RedisRestOptions
• ResourceBudget
• ResponseFormat
• Retriever
• RetryPolicy
• redactPostHook
• redis
• repairJson
• restoreFromSnapshot
• resumeFromHuman
• retryOnErrorPolicy
• runInputGuardrails
• runOutputGuardrails
• runToolGuardrails
• runToolPostHooks

S

• SCORE_JUDGE_SYSTEM_PROMPT
• ScalarLLMJudgeVerifier
• ScalarLLMJudgeVerifierOptions
• ScalarVerdict
• ScoutSnapshot
• SearchResult
• SetOptions
• StateDeltaOp
• Step
• StopCondition
• StopConditionContext
• StopPolicyDescriptor
• StreamEvent
• StreamingActionParser
• StructuredKvBackend
• StructuredMemory
• SubagentRunnable
• SummarizeOptions
• SupervisorAction
• SupervisorPolicy
• stepCountIs
• streaming
• summarizeToolOutput

T

• TfidfEmbedder
• TokenBudget
• TokenUsage
• ToolCall
• ToolCallingAgent
• ToolCallingAgentOptions
• ToolDefinition
• ToolGuardrail
• ToolGuardrailContext
• ToolPostHook
• ToolPostHookContext
• ToolRegistry
• ToolResult
• ToolUseStep
• toStrictJsonSchema
• tools
• truncatePostHook
• types

U

• UserMessageStep

V

• ValidationTask
• VerificationPipeline
• VerificationResult
• Verifier
• VisualAssertVerifier
• VisualAssertVerifierOptions
• VisualResult
• VisualResultReader
• VisualVerdict
• VmKernel

W

• WasmKernel
• WorkspaceReader
• WriteOpKind

Z

• zodToJsonSchema