Module rpc

Expand description

JSON-RPC request/response types and typed namespace builders for Client::rpc and session::Session::rpc. JSON-RPC request/response types and typed namespace builders.

All types are auto-generated from the Copilot CLI protocol schemas. This module is the stable public access point — the underlying crate-private modules where the types are defined are an implementation detail whose layout may change.

Use the crate::Client::rpc and crate::session::Session::rpc helper methods to obtain a typed view over the protocol surface.

Modules§

rpc_methods: JSON-RPC method name constants.

Structs§

AbortRequest: Parameters for aborting the current turn
AbortResult: Result of aborting the current turn
AccountAllUsers: Authenticated account entry returned by account.getAllUsers, with auth info and an optional associated token.
AccountGetCurrentAuthResult: Current authentication state
AccountGetQuotaRequest: Optional GitHub token used to look up quota for a specific user instead of the global auth context.
AccountGetQuotaResult: Quota usage snapshots for the resolved user, keyed by quota type.
AccountLoginRequest: Credentials to store after successful authentication
AccountLoginResult: Result of a successful login; throws on failure
AccountLogoutRequest: User to log out
AccountLogoutResult: Logout result indicating if more users remain
AccountQuotaSnapshot: Quota usage snapshot for a Copilot quota type, including entitlement, used requests, overage, reset date, and remaining percentage.
AgentDiscoveryPath: Canonical directory where custom agents can be discovered or created, with scope, preference, and optional project path.
AgentDiscoveryPathList: Canonical locations where custom agents can be created so the runtime will recognize them.
AgentGetCurrentResult: The currently selected custom agent, or null when using the default agent.
AgentInfo: Custom agent metadata, including identifiers, display details, source, tools, model, MCP servers, skills, and file path.
AgentList: Custom agents available to the session.
AgentRegistryLiveTargetEntry: Full registry entry for the spawned child. Lets the controller call handleLiveTargetSelected(entry) directly without re-reading the registry (avoids a TOCTOU window).
AgentRegistryLogCapture: Per-spawn log-capture outcome; populated from spawnLiveTarget.
AgentRegistrySpawnError: child_process.spawn itself failed before the child entered the registry.
AgentRegistrySpawnRegistryTimeout: Spawn succeeded but the child did not publish a matching managed-server entry within the timeout.
AgentRegistrySpawnRequest: Inputs to spawn a managed-server child via the controller’s spawn delegate.
AgentRegistrySpawnSpawned: Managed-server child was spawned and registered successfully.
AgentRegistrySpawnValidationError: Synchronous pre-validation rejected the spawn request.
AgentReloadResult: Custom agents available to the session after reloading definitions from disk.
AgentSelectRequest: Name of the custom agent to select for subsequent turns.
AgentSelectResult: The newly selected custom agent.
AgentsDiscoverRequest: Optional project paths to include in agent discovery.
AgentsDiscoverResult: Agents discovered across user, project, plugin, and remote sources.
AgentsGetDiscoveryPathsRequest: Optional project paths to include when enumerating agent discovery directories.
AgentsGetDiscoveryPathsResult: Canonical locations where custom agents can be created so the runtime will recognize them.
AllowAllPermissionSetResult: Indicates whether the operation succeeded and reports the post-mutation state.
AllowAllPermissionState: Current allow-all permission mode.
ApiKeyAuthInfo: Authentication-info variant for API-key authentication to a non-GitHub LLM provider, carrying the secret apiKey and host.
AttachmentBlob: Blob attachment with inline base64-encoded data
AttachmentDirectory: Directory attachment
AttachmentExtensionContext: Structured context contributed by an extension. Composer pills displayed in the host are forwarded back through session.send.attachments, then rendered into the model prompt as an <extension_context> XML block.
AttachmentFile: File attachment
AttachmentFileLineRange: Optional line range to scope the attachment to a specific section of the file
AttachmentGitHubActionsJob: Pointer to a GitHub Actions job.
AttachmentGitHubCommit: Pointer to a GitHub commit.
AttachmentGitHubFile: Pointer to a file in a GitHub repository at a specific ref.
AttachmentGitHubFileDiff: Pointer to a single-file diff. At least one of head and base must be present.
AttachmentGitHubFileDiffSide: One side of a file diff (head or base)
AttachmentGitHubReference: GitHub issue, pull request, or discussion reference
AttachmentGitHubRelease: Pointer to a GitHub release.
AttachmentGitHubRepository: Pointer to a GitHub repository.
AttachmentGitHubSnippet: Pointer to a line range inside a file in a GitHub repository.
AttachmentGitHubTreeComparison: Pointer to a comparison between two git revisions.
AttachmentGitHubTreeComparisonSide: One side of a tree comparison (head or base)
AttachmentGitHubUrl: Generic GitHub URL reference.
AttachmentSelection: Code selection attachment from an editor
AttachmentSelectionDetails: Position range of the selection within the file
AttachmentSelectionDetailsEnd: End position of the selection
AttachmentSelectionDetailsStart: Start position of the selection
CancelUserRequestedShellCommandResult: Cancellation result for a user-requested shell command.
CanvasAction: Canvas action that the agent or host can invoke. To discover the input schema for a particular action, call the list_canvas_capabilities tool.
CanvasActionInvokeRequest: Canvas action invocation parameters.
CanvasActionInvokeResult: Canvas action invocation result.
CanvasCloseRequest: Canvas close parameters.
CanvasHostContext: Host context supplied by the runtime.
CanvasHostContextCapabilities: Host capabilities
CanvasList: Declared canvases available in this session.
CanvasListOpenResult: Live open-canvas snapshot.
CanvasOpenRequest: Canvas open parameters.
CanvasOpenResult: Canvas open result returned by the provider.
CanvasProviderCloseRequest: Canvas close parameters sent to the provider.
CanvasProviderInvokeActionRequest: Canvas action invocation parameters sent to the provider.
CanvasProviderOpenRequest: Canvas open parameters sent to the provider.
CanvasProviderOpenResult: Canvas open result returned by the provider.
CanvasSessionContext: Session context supplied by the runtime.
CapiSessionOptions: Options scoped to the built-in CAPI (Copilot API) provider.
ClientRpc: Typed view over the Client’s server-level RPC namespace.
ClientRpcAccount: account.* RPCs.
ClientRpcAgentRegistry: agentRegistry.* RPCs.
ClientRpcAgents: agents.* RPCs.
ClientRpcInstructions: instructions.* RPCs.
ClientRpcLlmInference: llmInference.* RPCs.
ClientRpcMcp: mcp.* RPCs.
ClientRpcMcpConfig: mcp.config.* RPCs.
ClientRpcModels: models.* RPCs.
ClientRpcPlugins: plugins.* RPCs.
ClientRpcPluginsMarketplaces: plugins.marketplaces.* RPCs.
ClientRpcRuntime: runtime.* RPCs.
ClientRpcSecrets: secrets.* RPCs.
ClientRpcSessionFs: sessionFs.* RPCs.
ClientRpcSessions: sessions.* RPCs.
ClientRpcSkills: skills.* RPCs.
ClientRpcSkillsConfig: skills.config.* RPCs.
ClientRpcTools: tools.* RPCs.
ClientRpcUser: user.* RPCs.
ClientRpcUserSettings: user.settings.* RPCs.
CommandList: Slash commands available in the session, after applying any include/exclude filters.
CommandsHandlePendingCommandRequest: Pending command request ID and an optional error if the client handler failed.
CommandsHandlePendingCommandResult: Indicates whether the pending client-handled command was completed successfully.
CommandsInvokeRequest: Slash command name and optional raw input string to invoke.
CommandsListRequest: Optional filters controlling which command sources to include in the listing.
CommandsRespondToQueuedCommandRequest: Queued-command request ID and the result indicating whether the host executed it (and whether to stop processing further queued commands).
CommandsRespondToQueuedCommandResult: Indicates whether the queued-command response was matched to a pending request.
CompletionsGetTriggerCharactersResult: Characters that, when typed in the composer, should trigger a completions.request. Empty when the session has no host-driven completions (e.g. local sessions, or a relay host that does not advertise completionTriggerCharacters).
CompletionsRequestRequest: Request host-driven completions for the current composer input.
CompletionsRequestResult: Host-driven completion items for the current composer input. Empty when the host returns no items or does not support completions.
ConnectRemoteSessionParams: Remote session connection parameters.
ConnectedRemoteSessionMetadata: Metadata for a connected remote session.
ConnectedRemoteSessionMetadataRepository: Repository associated with the connected remote session.
ContextHeaviestMessage: A single large message currently in context.
CopilotApiTokenAuthInfo: Authentication-info variant for direct Copilot API token auth sourced from environment variables, with public GitHub host.
CopilotUserResponse: Snapshot of the authenticated user’s Copilot subscription info, if known. Mirrors the GitHub API /copilot_internal/v2/token user response shape — the runtime trusts this verbatim and does not re-fetch when set.
CopilotUserResponseEndpoints: Endpoint URLs from the raw Copilot /copilot_internal/v2/token user-response passthrough.
CopilotUserResponseQuotaSnapshots: Quota snapshot map from the raw Copilot user-response passthrough, with chat, completions, premium-interactions, and other entries.
CopilotUserResponseQuotaSnapshotsChat: Chat quota snapshot from the raw Copilot user-response passthrough, with entitlement, overage, remaining quota, reset, and billing fields.
CopilotUserResponseQuotaSnapshotsCompletions: Completions quota snapshot from the raw Copilot user-response passthrough, with entitlement, overage, remaining quota, reset, and billing fields.
CopilotUserResponseQuotaSnapshotsPremiumInteractions: Premium-interactions quota snapshot from the raw Copilot user-response passthrough, with entitlement, overage, remaining quota, reset, and billing fields.
CurrentModel: The currently selected model, reasoning effort, and context tier for the session. The context tier reflects Session.getContextTier(), restored from the session journal on resume.
CurrentToolMetadata: Lightweight metadata for a currently initialized session tool
DebugCollectLogsCollectedEntry: A file included in the redacted debug bundle.
DebugCollectLogsDestinationArchive
DebugCollectLogsDestinationDirectory
DebugCollectLogsEntry: A caller-provided server-local file or directory to include in the debug bundle.
DebugCollectLogsInclude: Built-in session diagnostics to include in the bundle. Omitted fields default to true.
DebugCollectLogsRequest: Options for collecting a redacted session debug bundle.
DebugCollectLogsResult: Result of collecting a redacted debug bundle.
DebugCollectLogsSkippedEntry: An optional debug bundle entry that could not be included.
DiscoveredCanvas: Canvas available in the current session.
DiscoveredMcpServer: MCP server discovered by mcp.discover, with config source, optional plugin source, transport type, and enabled state.
EnqueueCommandParams: Slash-prefixed command string to enqueue for FIFO processing.
EnqueueCommandResult: Indicates whether the command was accepted into the local execution queue.
EnvAuthInfo: Authentication-info variant for a token sourced from an environment variable, with host, optional login, token, and env var name.
EventLogReadRequest: Cursor, batch size, and optional long-poll/filter parameters for reading session events.
EventLogReleaseInterestResult: Indicates whether the operation succeeded.
EventLogTailResult: Snapshot of the current tail cursor without returning any events. Use this when a consumer wants to subscribe to live events going forward without first paginating through the entire persisted history (which would happen if read were called without a cursor on a long-lived session).
EventsReadResult: Batch of session events returned by a read, with cursor and continuation metadata.
ExecuteCommandParams: Slash command name and argument string to execute synchronously.
ExecuteCommandResult: Error message produced while executing the command, if any.
Extension: Discovered extension metadata, including source-qualified ID, name, discovery source, status, and optional process ID.
ExtensionContextPushInput: Slim input shape for extension_context attachments; identity fields are runtime-derived.
ExtensionList: Extensions discovered for the session, with their current status.
ExtensionsDisableRequest: Source-qualified extension identifier to disable for the session.
ExtensionsEnableRequest: Source-qualified extension identifier to enable for the session.
ExternalToolTextResultForLlm: Expanded external tool result payload
ExternalToolTextResultForLlmBinaryResultsForLlm: Binary result returned by a tool for the model
ExternalToolTextResultForLlmContentAudio: Audio content block with base64-encoded data
ExternalToolTextResultForLlmContentImage: Image content block with base64-encoded data
ExternalToolTextResultForLlmContentResource: Embedded resource content block with inline text or binary data
ExternalToolTextResultForLlmContentResourceLink: Resource link content block referencing an external resource
ExternalToolTextResultForLlmContentResourceLinkIcon: Icon image for a resource
ExternalToolTextResultForLlmContentShellExit: Shell command exit metadata with optional output preview
ExternalToolTextResultForLlmContentTerminal: Terminal/shell output content block with optional exit code and working directory
ExternalToolTextResultForLlmContentText: Plain text content block
FleetStartRequest: Optional user prompt to combine with the fleet orchestration instructions.
FleetStartResult: Indicates whether fleet mode was successfully activated.
FolderTrustAddParams: Folder path to add to trusted folders.
FolderTrustCheckParams: Folder path to check for trust.
FolderTrustCheckResult: Folder trust check result.
GhCliAuthInfo: Authentication-info variant for GitHub CLI credentials, carrying host, login, and the gh auth token value.
GitHubRepoRef: Pointer to a GitHub repository.
GitHubTelemetryClientInfo: Client environment metadata describing the process that produced a telemetry event.
GitHubTelemetryEvent: A single telemetry event in the runtime’s native GitHub-shaped telemetry format, forwarded verbatim to opted-in hosts. The restricted flag on the enclosing GitHubTelemetryNotification distinguishes standard from restricted events; the payload shape is identical for both.
GitHubTelemetryNotification: Payload for a gitHubTelemetry.event notification: a single GitHub telemetry event the runtime forwards to a host connection that opted into telemetry forwarding during the server.connect handshake.
HMACAuthInfo: Authentication-info variant for GitHub-internal HMAC auth, carrying the public GitHub host and HMAC secret.
HandlePendingToolCallRequest: Pending external tool call request ID, with the tool result or an error describing why it failed.
HandlePendingToolCallResult: Indicates whether the external tool call result was handled successfully.
HistoryAbortManualCompactionResult: Indicates whether an in-progress manual compaction was aborted.
HistoryCancelBackgroundCompactionResult: Indicates whether an in-progress background compaction was cancelled.
HistoryCompactContextWindow: Post-compaction context window usage breakdown
HistoryCompactRequest: Optional compaction parameters.
HistoryCompactResult: Compaction outcome with the number of tokens and messages removed, summary text, and the resulting context window breakdown.
HistorySummarizeForHandoffResult: Markdown summary of the conversation context (empty when not available).
HistoryTruncateRequest: Identifier of the event to truncate to; this event and all later events are removed.
HistoryTruncateResult: Number of events that were removed by the truncation.
InstalledPlugin: Installed plugin record from global state, with marketplace, version, install time, enabled state, cache path, and source.
InstalledPluginInfo: Information about an installed plugin tracked in global state.
InstalledPluginSourceGitHub: Source descriptor for a direct GitHub plugin install, with owner/repo, optional ref, and optional subpath.
InstalledPluginSourceLocal: Source descriptor for a direct local plugin install, with a local filesystem path.
InstalledPluginSourceUrl: Source descriptor for a direct URL plugin install, with URL, optional ref, and optional subpath.
InstructionDiscoveryPath: Canonical file or directory where custom instructions can be discovered or created, with location, kind, preference, and project path.
InstructionDiscoveryPathList: Canonical files and directories where custom instructions can be created so the runtime will recognize them.
InstructionSource: Loaded instruction source for a session, including path, content, category, location, applicability, and optional description.
InstructionsDiscoverRequest: Optional project paths to include in instruction discovery.
InstructionsDiscoverResult: Instruction sources discovered across user, repository, and plugin sources.
InstructionsGetDiscoveryPathsRequest: Optional project paths to include when enumerating instruction discovery targets.
InstructionsGetDiscoveryPathsResult: Canonical files and directories where custom instructions can be created so the runtime will recognize them.
InstructionsGetSourcesResult: Instruction sources loaded for the session, in merge order.
LlmInferenceHttpRequestChunkRequest: A request body chunk or cancellation signal.
LlmInferenceHttpRequestChunkResult: Acknowledgement. The SDK is free to ignore the ack and treat chunk delivery as fire-and-forget.
LlmInferenceHttpRequestStartRequest: The head of an outbound model-layer HTTP request.
LlmInferenceHttpRequestStartResult: Acknowledgement. Returning successfully simply means the SDK accepted the start frame; it does not imply the request will succeed.
LlmInferenceHttpResponseChunkError: Set to terminate the response with a transport-level failure. Implies end-of-stream; any further chunks for this requestId are ignored.
LlmInferenceHttpResponseChunkRequest: A response body chunk or terminal error.
LlmInferenceHttpResponseChunkResult: Whether the chunk was accepted.
LlmInferenceHttpResponseStartRequest: Response head.
LlmInferenceHttpResponseStartResult: Whether the start frame was accepted.
LlmInferenceSetProviderResult: Indicates whether the calling client was registered as the LLM inference provider.
LocalSessionMetadataValue: Persisted local session metadata, including identifiers, timestamps, summary/name, client, context, detached state, and task ID.
LogRequest: Message text, optional severity level, persistence flag, optional follow-up URL, and optional tip.
LogResult: Identifier of the session event that was emitted for the log message.
LspInitializeRequest: Parameters for (re)loading the merged LSP configuration set.
MarketplaceAddResult: Result of registering a new marketplace.
MarketplaceBrowseResult: Plugins advertised by the marketplace.
MarketplaceInfo: Registered marketplace summary.
MarketplaceListResult: All registered marketplaces, including built-in defaults.
MarketplacePluginInfo: Plugin entry advertised by a marketplace.
MarketplaceRefreshEntry: Per-marketplace refresh result, including marketplace name, success flag, and optional failure error.
MarketplaceRefreshResult: Result of refreshing one or more marketplace catalogs.
MarketplaceRemoveResult: Outcome of the remove attempt, including dependent-plugin info when applicable.
McpAllowedServer: MCP server allowed by policy, with server name and optional PII-free explanatory note.
McpAppsCallToolRequest: MCP server, tool name, and arguments to invoke from an MCP App view.
McpAppsDiagnoseCapability: Capability negotiation snapshot
McpAppsDiagnoseRequest: MCP server to diagnose MCP Apps wiring for.
McpAppsDiagnoseResult: Diagnostic snapshot of MCP Apps wiring for the named server.
McpAppsDiagnoseServer: What the server returned for this session
McpAppsHostContext: Current host context advertised to MCP App guests.
McpAppsHostContextDetails: Current host context
McpAppsListToolsRequest: MCP server to list app-callable tools for.
McpAppsListToolsResult: App-callable tools from the named MCP server.
McpAppsReadResourceRequest: MCP server and resource URI to fetch.
McpAppsReadResourceResult: Resource contents returned by the MCP server.
McpAppsResourceContent: MCP Apps resource content with URI, optional MIME type, text or base64 blob, and resource metadata.
McpAppsSetHostContextDetails: Host context advertised to MCP App guests
McpAppsSetHostContextRequest: Host context to advertise to MCP App guests.
McpCancelSamplingExecutionParams: The requestId previously passed to executeSampling that should be cancelled.
McpCancelSamplingExecutionResult: Indicates whether an in-flight sampling execution with the given requestId was found and cancelled.
McpConfigAddRequest: MCP server name and configuration to add to user configuration.
McpConfigDisableRequest: MCP server names to disable for new sessions.
McpConfigEnableRequest: MCP server names to enable for new sessions.
McpConfigList: User-configured MCP servers, keyed by server name.
McpConfigListResult: User-configured MCP servers, keyed by server name.
McpConfigRemoveRequest: MCP server name to remove from user configuration.
McpConfigUpdateRequest: MCP server name and replacement configuration to write to user configuration.
McpConfigureGitHubResult: Result of configuring GitHub MCP.
McpDisableRequest: Name of the MCP server to disable for the session.
McpDiscoverRequest: Optional working directory used as context for MCP server discovery.
McpDiscoverResult: MCP servers discovered from user, workspace, plugin, and built-in sources.
McpEnableRequest: Name of the MCP server to enable for the session.
McpExecuteSamplingParams: Identifiers and raw MCP CreateMessageRequest params used to run a sampling inference.
McpExecuteSamplingRequest: Raw MCP CreateMessageRequest params, as received in the sampling.requested event. Treated as opaque at the schema layer; the runtime converts the embedded MCP messages into the OpenAI chat-completion shape internally.
McpFilteredServer: MCP server filtered by policy, with name, reason, optional redacted reason, and enterprise login.
McpHeadersHandlePendingHeadersRefreshRequestHeaders
McpHeadersHandlePendingHeadersRefreshRequestNone
McpHeadersHandlePendingHeadersRefreshRequestRequest: MCP headers refresh request id and the host response.
McpHeadersHandlePendingHeadersRefreshRequestResult: Indicates whether the pending MCP headers refresh response was accepted.
McpHostState: Host-level state, omitted when no MCP host is initialized.
McpIsServerRunningRequest: Server name to check running status for.
McpIsServerRunningResult: Whether the named MCP server is running.
McpListToolsRequest: Server name whose tool list should be returned.
McpListToolsResult: Tools exposed by the connected MCP server. Throws when the server is not connected.
McpOauthHandlePendingRequest: Pending MCP OAuth request ID and host-provided token or cancellation response.
McpOauthHandlePendingResult: Indicates whether the pending MCP OAuth response was accepted.
McpOauthLoginRequest: Remote MCP server name and optional overrides controlling reauthentication, OAuth client display name, callback success-page copy, and static OAuth client selection.
McpOauthLoginResult: OAuth authorization URL the caller should open, or empty when cached tokens already authenticated the server.
McpOauthPendingRequestResponseCancelled
McpOauthPendingRequestResponseToken
McpRemoveGitHubResult: Indicates whether the auto-managed github MCP server was removed (false when nothing to remove).
McpSamplingExecutionResult: Outcome of an MCP sampling execution: success result, failure error, or cancellation.
McpServer: MCP server status entry, including config source/plugin source and any connection error.
McpServerAuthConfigRedirectPort: Authentication settings with optional redirect port configuration.
McpServerConfigHttp: Remote MCP server configuration accessed over HTTP or SSE.
McpServerConfigStdio: Stdio MCP server configuration launched as a child process.
McpServerFailureInfo: Recorded MCP server connection failure.
McpServerList: MCP servers configured for the session, with their connection status and host-level state.
McpServerNeedsAuthInfo: Recorded MCP server pending-auth state.
McpSetEnvValueModeParams: Mode controlling how MCP server env values are resolved (direct or indirect).
McpSetEnvValueModeResult: Env-value mode recorded on the session after the update.
McpStartServersResult: MCP server startup filtering result.
McpStopServerRequest: Server name for an individual MCP server stop.
McpTools: MCP tool metadata with tool name and optional description.
MemoryConfiguration: Memory configuration for this session.
MetadataContextAttributionResult: Per-source attribution breakdown for the session’s current context window, or null if uninitialized.
MetadataContextAttributionResultContextAttribution: Per-source token attribution snapshot for the current context window. The heaviest individual messages are available separately via metadata.getContextHeaviestMessages.
MetadataContextAttributionResultContextAttributionCompactions: Successful compaction history for the session.
MetadataContextAttributionResultContextAttributionEntriesItem
MetadataContextHeaviestMessagesRequest: Parameters for the heaviest-messages query.
MetadataContextHeaviestMessagesResult: The heaviest individual messages in the session’s context window, most-expensive first.
MetadataContextInfoRequest: Model identifier and token limits used to compute the context-info breakdown.
MetadataContextInfoResult: Token breakdown for the session’s current context window, or null if uninitialized.
MetadataContextInfoResultContextInfo: Token-usage breakdown for the session’s current context window
MetadataIsProcessingResult: Indicates whether the local session is currently processing a turn or background continuation.
MetadataRecomputeContextTokensRequest: Model identifier to use when re-tokenizing the session’s existing messages.
MetadataRecomputeContextTokensResult: Re-tokenize the session’s existing messages against modelId and return the token totals. Useful for hosts that want an initial estimate of context usage on session resume, before the next agent turn fires session.context_info_changed events. Returns zeros for an empty session.
MetadataRecordContextChangeRequest: Updated working-directory/git context to record on the session.
MetadataRecordContextChangeResult: Notify the session that its working directory context has changed. Emits a session.context_changed event so consumers (telemetry, OTel tracker, ACP, the timeline UI) can react. Use this when the host has detected a cwd/branch/repo change outside the session’s normal lifecycle (e.g., after a shell command in interactive mode).
MetadataSetWorkingDirectoryRequest: Absolute path to set as the session’s new working directory.
MetadataSetWorkingDirectoryResult: Update the session’s working directory. Used by the host when the user explicitly changes cwd (e.g., the /cd slash command). The host is responsible for process.chdir and any related side-effects (file index, etc.); this method only updates the session’s own recorded path.
MetadataSnapshotRemoteMetadata: Remote-session-specific metadata. Populated only when isRemote is true. Fields are immutable for the lifetime of the session.
MetadataSnapshotRemoteMetadataRepository: The repository the remote session targets.
ModeSetRequest: Agent interaction mode to apply to the session.
Model: Copilot model metadata, including identifier, display name, capabilities, policy, billing, reasoning efforts, and picker categories.
ModelBilling: Billing information
ModelBillingTokenPrices: Token-level pricing information for this model
ModelBillingTokenPricesLongContext: Long context tier pricing (available for models with extended context windows)
ModelCapabilities: Model capabilities and limits
ModelCapabilitiesLimits: Token limits for prompts, outputs, and context window
ModelCapabilitiesLimitsVision: Vision-specific limits
ModelCapabilitiesOverride: Optional capability overrides (vision, tool_calls, reasoning, etc.).
ModelCapabilitiesOverrideLimits: Token limits for prompts, outputs, and context window
ModelCapabilitiesOverrideLimitsVision: Vision-specific limits
ModelCapabilitiesOverrideSupports: Feature flags indicating what the model supports
ModelCapabilitiesSupports: Feature flags indicating what the model supports
ModelList: List of Copilot models available to the resolved user, including capabilities and billing metadata.
ModelListRequest: Optional listing options.
ModelPolicy: Policy state (if applicable)
ModelSetReasoningEffortRequest: Reasoning effort level to apply to the currently selected model.
ModelSetReasoningEffortResult: Update the session’s reasoning effort without changing the selected model. Use switchTo instead when you also need to change the model. The runtime stores the effort on the session and applies it to subsequent turns.
ModelSwitchToRequest: Target model identifier and optional reasoning effort, summary, capability overrides, and context tier.
ModelSwitchToResult: The model identifier active on the session after the switch.
ModelsListRequest: Optional GitHub token used to list models for a specific user instead of the global auth context.
ModelsListResult: List of Copilot models available to the resolved user, including capabilities and billing metadata.
NameGetResult: The session’s friendly name, or null when not yet set.
NameSetAutoRequest: Auto-generated session summary to apply as the session’s name when no user-set name exists.
NameSetAutoResult: Indicates whether the auto-generated summary was applied as the session’s name.
NameSetRequest: New friendly name to apply to the session.
NamedProviderConfig: A named BYOK provider connection (transport + credentials).
OpenCanvasInstance: Open canvas instance snapshot.
OptionsUpdateAdditionalContentExclusionPolicy: Content-exclusion policy supplied to session.options.update, with rules, last-updated data, and scope.
OptionsUpdateAdditionalContentExclusionPolicyRule: Single content-exclusion rule supplied to session.options.update, with paths, match conditions, and source.
OptionsUpdateAdditionalContentExclusionPolicyRuleSource: Source descriptor for a session.options.update content-exclusion rule, with source name and type.
PendingPermissionRequest: Pending permission prompt reconstructed from event history, with request ID and user-facing prompt details.
PendingPermissionRequestList: List of pending permission requests reconstructed from event history.
PermissionDecisionApproveForLocation: Permission-decision request variant to approve and persist a permission for a project location, with approval details and location key.
PermissionDecisionApproveForLocationApprovalCommands: Location-scoped approval details for specific command identifiers.
PermissionDecisionApproveForLocationApprovalCustomTool: Location-scoped approval details for a custom tool, keyed by tool name.
PermissionDecisionApproveForLocationApprovalExtensionManagement: Location-scoped approval details for extension-management operations, optionally narrowed by operation.
PermissionDecisionApproveForLocationApprovalExtensionPermissionAccess: Location-scoped approval details for an extension’s permission-gated capability access, keyed by extension name.
PermissionDecisionApproveForLocationApprovalMcp: Location-scoped approval details for an MCP server tool, or all tools on the server when toolName is null.
PermissionDecisionApproveForLocationApprovalMcpSampling: Location-scoped approval details for MCP sampling requests from a server.
PermissionDecisionApproveForLocationApprovalMemory: Location-scoped approval details for writes to long-term memory.
PermissionDecisionApproveForLocationApprovalRead: Location-scoped approval details for read-only filesystem operations.
PermissionDecisionApproveForLocationApprovalWrite: Location-scoped approval details for filesystem write operations.
PermissionDecisionApproveForSession: Permission-decision request variant to approve for the rest of the session, with optional tool approval or URL domain.
PermissionDecisionApproveForSessionApprovalCommands: Session-scoped approval details for specific command identifiers.
PermissionDecisionApproveForSessionApprovalCustomTool: Session-scoped approval details for a custom tool, keyed by tool name.
PermissionDecisionApproveForSessionApprovalExtensionManagement: Session-scoped approval details for extension-management operations, optionally narrowed by operation.
PermissionDecisionApproveForSessionApprovalExtensionPermissionAccess: Session-scoped approval details for an extension’s permission-gated capability access, keyed by extension name.
PermissionDecisionApproveForSessionApprovalMcp: Session-scoped approval details for an MCP server tool, or all tools on the server when toolName is null.
PermissionDecisionApproveForSessionApprovalMcpSampling: Session-scoped approval details for MCP sampling requests from a server.
PermissionDecisionApproveForSessionApprovalMemory: Session-scoped approval details for writes to long-term memory.
PermissionDecisionApproveForSessionApprovalRead: Session-scoped approval details for read-only filesystem operations.
PermissionDecisionApproveForSessionApprovalWrite: Session-scoped approval details for filesystem write operations.
PermissionDecisionApproveOnce: Permission-decision request variant to approve only the current permission request.
PermissionDecisionApprovePermanently: Permission-decision request variant to permanently approve a URL domain across sessions.
PermissionDecisionApproved: Permission-decision variant indicating the request was approved.
PermissionDecisionApprovedForLocation: Permission-decision variant indicating approval was persisted for a project location, with approval details and location key.
PermissionDecisionApprovedForSession: Permission-decision variant indicating approval was remembered for the session, with approval details.
PermissionDecisionCancelled: Permission-decision variant indicating the request was cancelled before use, with an optional reason.
PermissionDecisionDeniedByContentExclusionPolicy: Permission-decision variant indicating denial by content-exclusion policy, with path and message.
PermissionDecisionDeniedByPermissionRequestHook: Permission-decision variant indicating denial by a permission request hook, with optional message and interrupt flag.
PermissionDecisionDeniedByRules: Permission-decision variant indicating explicit denial by permission rules, with the matching rules.
PermissionDecisionDeniedInteractivelyByUser: Permission-decision variant indicating the user denied an interactive prompt, with optional feedback and force-reject flag.
PermissionDecisionDeniedNoApprovalRuleAndCouldNotRequestFromUser: Permission-decision variant indicating no approval rule matched and user confirmation was unavailable.
PermissionDecisionReject: Permission-decision request variant to reject a pending permission request, with optional feedback.
PermissionDecisionRequest: Pending permission request ID and the decision to apply (approve/reject and scope).
PermissionDecisionUserNotAvailable: Permission-decision variant indicating no user was available to confirm the request.
PermissionLocationAddToolApprovalParams: Location-scoped tool approval to persist.
PermissionLocationApplyParams: Working directory to load persisted location permissions for.
PermissionLocationApplyResult: Summary of persisted location permissions applied to the session.
PermissionLocationResolveParams: Working directory to resolve into a location-permissions key.
PermissionLocationResolveResult: Resolved location-permissions key and type.
PermissionPathsAddParams: Directory path to add to the session’s allowed directories.
PermissionPathsAllowedCheckParams: Path to evaluate against the session’s allowed directories.
PermissionPathsAllowedCheckResult: Indicates whether the supplied path is within the session’s allowed directories.
PermissionPathsConfig: If specified, replaces the session’s path-permission policy. The runtime constructs the appropriate PathManager based on these inputs (rooted at the session’s working directory). Omit to leave the current path policy unchanged.
PermissionPathsList: Snapshot of the session’s allow-listed directories and primary working directory.
PermissionPathsUpdatePrimaryParams: Directory path to set as the session’s new primary working directory.
PermissionPathsWorkspaceCheckParams: Path to evaluate against the session’s workspace (primary) directory.
PermissionPathsWorkspaceCheckResult: Indicates whether the supplied path is within the session’s workspace directory.
PermissionPromptShownNotification: Notification payload describing the permission prompt that the client just rendered.
PermissionRequestResult: Indicates whether the permission decision was applied; false when the request was already resolved.
PermissionRulesSet: If specified, replaces the session’s approved/denied permission rules. Omit to leave the current rules unchanged.
PermissionUrlsConfig: If specified, replaces the session’s URL-permission policy. The runtime constructs a fresh DefaultUrlManager based on these inputs. Omit to leave the current URL policy unchanged.
PermissionUrlsSetUnrestrictedModeParams: Whether the URL-permission policy should run in unrestricted mode.
PermissionsConfigureAdditionalContentExclusionPolicy: Content-exclusion policy supplied to session.permissions.configure, with rules, last-updated data, and scope.
PermissionsConfigureAdditionalContentExclusionPolicyRule: Single content-exclusion rule supplied to session.permissions.configure, with paths, match conditions, and source.
PermissionsConfigureAdditionalContentExclusionPolicyRuleSource: Source descriptor for a session.permissions.configure content-exclusion rule, with source name and type.
PermissionsConfigureParams: Patch of permission policy fields to apply (omit a field to leave it unchanged).
PermissionsConfigureResult: Indicates whether the operation succeeded.
PermissionsFolderTrustAddTrustedResult: Indicates whether the operation succeeded.
PermissionsGetAllowAllRequest: No parameters.
PermissionsLocationsAddToolApprovalDetailsCommands: Location-persisted tool approval details for specific command identifiers.
PermissionsLocationsAddToolApprovalDetailsCustomTool: Location-persisted tool approval details for a custom tool, keyed by tool name.
PermissionsLocationsAddToolApprovalDetailsExtensionManagement: Location-persisted tool approval details for extension-management operations, optionally narrowed by operation.
PermissionsLocationsAddToolApprovalDetailsExtensionPermissionAccess: Location-persisted tool approval details for an extension’s permission-gated capability access, keyed by extension name.
PermissionsLocationsAddToolApprovalDetailsMcp: Location-persisted tool approval details for an MCP server tool, or all tools when toolName is null.
PermissionsLocationsAddToolApprovalDetailsMcpSampling: Location-persisted tool approval details for MCP sampling requests from a server.
PermissionsLocationsAddToolApprovalDetailsMemory: Location-persisted tool approval details for writes to long-term memory.
PermissionsLocationsAddToolApprovalDetailsRead: Location-persisted tool approval details for read-only filesystem operations.
PermissionsLocationsAddToolApprovalDetailsWrite: Location-persisted tool approval details for filesystem write operations.
PermissionsLocationsAddToolApprovalResult: Indicates whether the operation succeeded.
PermissionsModifyRulesParams: Scope and add/remove instructions for modifying session- or location-scoped permission rules.
PermissionsModifyRulesResult: Indicates whether the operation succeeded.
PermissionsNotifyPromptShownResult: Indicates whether the operation succeeded.
PermissionsPathsAddResult: Indicates whether the operation succeeded.
PermissionsPathsListRequest: No parameters; returns the session’s allow-listed directories.
PermissionsPathsUpdatePrimaryResult: Indicates whether the operation succeeded.
PermissionsPendingRequestsRequest: No parameters; returns currently-pending permission requests for the session.
PermissionsResetSessionApprovalsRequest: No parameters; clears all session-scoped tool permission approvals.
PermissionsResetSessionApprovalsResult: Indicates whether the operation succeeded.
PermissionsSetAllowAllRequest: Allow-all mode to apply for the session.
PermissionsSetApproveAllRequest: Allow-all toggle for tool permission requests, with an optional telemetry source.
PermissionsSetApproveAllResult: Indicates whether the operation succeeded.
PermissionsSetRequiredRequest: Toggles whether permission prompts should be bridged into session events for this client.
PermissionsSetRequiredResult: Indicates whether the operation succeeded.
PermissionsUrlsSetUnrestrictedModeResult: Indicates whether the operation succeeded.
PingRequest: Optional message to echo back to the caller.
PingResult: Server liveness response, including the echoed message, current server timestamp, and protocol version.
PlanReadResult: Existence, contents, and resolved path of the session plan file.
PlanReadSqlTodosResult: Todo rows read from the session SQL database. Empty when no session database is available.
PlanReadSqlTodosWithDependenciesResult: Todo rows + dependency edges read from the session SQL database.
PlanSqlTodoDependency: A single dependency edge read from the session SQL todo_deps table, indicating that one todo must complete before another.
PlanSqlTodosRow: A single todo row read from the session SQL todos table. All fields are optional because the SQL schema is best-effort and the agent may not have populated every column.
PlanUpdateRequest: Replacement contents to write to the session plan file.
Plugin: Session plugin metadata, with name, marketplace, optional version, and enabled state.
PluginInstallResult: Result of installing a plugin.
PluginList: Plugins installed for the session, with their enabled state and version metadata.
PluginListResult: Plugins installed in user/global state.
PluginUpdateAllEntry: Per-plugin result from updating all plugins, with versions, skills installed, success flag, and optional error.
PluginUpdateAllResult: Result of updating all installed plugins.
PluginUpdateResult: Result of updating a single plugin.
PluginsDisableRequest: Plugin names (or specs) to disable.
PluginsEnableRequest: Plugin names (or specs) to enable.
PluginsInstallRequest: Plugin source and optional working directory for relative-path resolution.
PluginsInstallResult: Result of installing a plugin.
PluginsListResult: Plugins installed in user/global state.
PluginsMarketplacesAddRequest: Marketplace source to register.
PluginsMarketplacesAddResult: Result of registering a new marketplace.
PluginsMarketplacesBrowseRequest: Name of the marketplace whose plugin catalog to fetch.
PluginsMarketplacesBrowseResult: Plugins advertised by the marketplace.
PluginsMarketplacesListResult: All registered marketplaces, including built-in defaults.
PluginsMarketplacesRefreshRequest: Optional marketplace name; omit to refresh all.
PluginsMarketplacesRefreshResult: Result of refreshing one or more marketplace catalogs.
PluginsMarketplacesRemoveRequest: Name of the marketplace to remove and an optional force flag.
PluginsMarketplacesRemoveResult: Outcome of the remove attempt, including dependent-plugin info when applicable.
PluginsReloadRequest: Optional flags controlling which side effects the reload performs.
PluginsUninstallRequest: Name (or spec) of the plugin to uninstall.
PluginsUpdateAllResult: Result of updating all installed plugins.
PluginsUpdateRequest: Name (or spec) of the plugin to update.
PluginsUpdateResult: Result of updating a single plugin.
ProviderAddRequest: BYOK providers and/or models to add to the session’s registry at runtime. Both fields are optional; provide providers, models, or both.
ProviderAddResult: The selectable model entries synthesized for the models added by this call.
ProviderConfig: Custom model-provider configuration (BYOK).
ProviderConfigAzure: Azure-specific provider options.
ProviderEndpoint: A snapshot of the provider endpoint the session is currently configured to talk to.
ProviderGetEndpointRequest: Optional model identifier to scope the endpoint snapshot to.
ProviderModelConfig: A BYOK model definition referencing a named provider.
ProviderSessionToken: Short-lived, rotating credential the caller must send on every request, in addition to apiKey if one is present. Omitted when the endpoint does not require one.
ProviderTokenAcquireRequest: Asks the SDK client to acquire a bearer token for a BYOK provider whose config set hasBearerTokenProvider: true. Issued by the runtime before each outbound model request; the runtime does no caching, so this is sent once per request.
ProviderTokenAcquireResult: A bearer token supplied by the SDK client for a BYOK provider. The runtime sets it as Authorization: Bearer <token> on the outbound request and does no caching; the SDK consumer owns token caching and refresh.
ProviderTokenGetTokenResult: A bearer token supplied by the SDK client for a BYOK provider. The runtime sets it as Authorization: Bearer <token> on the outbound request and does no caching; the SDK consumer owns token caching and refresh.
PushAttachmentBlob: Blob attachment with inline base64-encoded data
PushAttachmentDirectory: Directory attachment
PushAttachmentFile: File attachment
PushAttachmentFileLineRange: Optional line range to scope the attachment to a specific section of the file
PushAttachmentGitHubActionsJob: Pointer to a GitHub Actions job.
PushAttachmentGitHubCommit: Pointer to a GitHub commit.
PushAttachmentGitHubFile: Pointer to a file in a GitHub repository at a specific ref.
PushAttachmentGitHubFileDiff: Pointer to a single-file diff. At least one of head and base must be present.
PushAttachmentGitHubFileDiffSide: One side of a file diff (head or base)
PushAttachmentGitHubReference: GitHub issue, pull request, or discussion reference
PushAttachmentGitHubRelease: Pointer to a GitHub release.
PushAttachmentGitHubRepository: Pointer to a GitHub repository.
PushAttachmentGitHubSnippet: Pointer to a line range inside a file in a GitHub repository.
PushAttachmentGitHubTreeComparison: Pointer to a comparison between two git revisions.
PushAttachmentGitHubTreeComparisonSide: One side of a tree comparison (head or base)
PushAttachmentGitHubUrl: Generic GitHub URL reference.
PushAttachmentSelection: Code selection attachment from an editor
PushAttachmentSelectionDetails: Position range of the selection within the file
PushAttachmentSelectionDetailsEnd: End position of the selection
PushAttachmentSelectionDetailsStart: Start position of the selection
PushGitHubRepoRef: Pointer to a GitHub repository.
QueuePendingItems: User-facing pending queue entry, with kind and display text for a queued message, slash command, or model change.
QueuePendingItemsResult: Snapshot of the session’s pending queued items and immediate-steering messages.
QueueRemoveMostRecentResult: Indicates whether a user-facing pending item was removed.
QueuedCommandHandled: Queued-command response indicating the host executed the command, with an optional flag to stop queue processing.
QueuedCommandNotHandled: Queued-command response indicating the host did not execute the command and the queue may continue.
RegisterEventInterestParams: Event type to register consumer interest for, used by runtime gating logic.
RegisterEventInterestResult: Opaque handle representing an event-type interest registration.
ReleaseEventInterestParams: Opaque handle previously returned by registerInterest to release.
RemoteControlConfig: Configuration for the runtime-managed remote-control singleton.
RemoteControlConfigExistingMcSession: Reattach to an existing MC session without creating a new one.
RemoteControlStatusActive: Remote control is connected to a local session.
RemoteControlStatusConnecting: Remote control is in the middle of initial setup.
RemoteControlStatusError: The last setup attempt failed. The singleton is otherwise off.
RemoteControlStatusOff: Remote control is not connected.
RemoteControlStatusResult: Wrapper for the singleton’s current status.
RemoteControlStopResult: Outcome of a stopRemoteControl call.
RemoteControlTransferResult: Outcome of a transferRemoteControl call.
RemoteEnableRequest: Optional remote session mode (“off”, “export”, or “on”); defaults to enabling both export and remote steering.
RemoteEnableResult: GitHub URL for the session and a flag indicating whether remote steering is enabled.
RemoteNotifySteerableChangedRequest: New remote-steerability state to persist as a session.remote_steerable_changed event.
RemoteNotifySteerableChangedResult: Persist a steerability change as a session.remote_steerable_changed event. Used by the host (CLI / SDK consumer) when it has just finished enabling or disabling steering on a remote exporter that the runtime does not directly own.
RemoteSessionConnectionResult: Remote session connection result.
RemoteSessionMetadataRepository: GitHub repository the remote session belongs to.
RemoteSessionMetadataValue: Remote session metadata for the session to hand off (typically obtained from sessions.list with source: "remote").
RemoteSessionRepository: Repository context for the remote session.
SandboxConfig: Resolved sandbox configuration.
SandboxConfigUserPolicy: User-managed sandbox policy fragment merged into the auto-discovered base policy.
SandboxConfigUserPolicyExperimental: Platform-specific experimental policy fields.
SandboxConfigUserPolicyExperimentalSeatbelt: macOS seatbelt experimental options.
SandboxConfigUserPolicyFilesystem: Filesystem rules to merge into the base policy.
SandboxConfigUserPolicyNetwork: Network rules to merge into the base policy.
SandboxConfigUserPolicySeatbelt: macOS seatbelt-specific options.
ScheduleEntry: Scheduled prompt entry with ID, timing (intervalMs, cron, or at), prompt text, recurrence, and next run time.
ScheduleList: Snapshot of the currently active recurring prompts for this session.
ScheduleStopRequest: Identifier of the scheduled prompt to remove.
ScheduleStopResult: Remove a scheduled prompt by id. The result entry is omitted if the id was unknown.
SecretsAddFilterValuesRequest: Secret values to add to the redaction filter.
SecretsAddFilterValuesResult: Confirmation that the secret values were registered.
SendAttachmentsToMessageParams: Parameters for session.extensions.sendAttachmentsToMessage.
SendRequest: Parameters for sending a user message to the session
SendResult: Result of sending a user message
ServerAgentList: Agents discovered across user, project, plugin, and remote sources.
ServerInstructionSourceList: Instruction sources discovered across user, repository, and plugin sources.
ServerSkill: Server-side skill metadata, including name, description, source, enabled/invocable state, path, project path, and argument hint.
ServerSkillList: Skills discovered across global and project sources.
SessionAbortResult: Result of aborting the current turn
SessionActivity: Current activity flags for the session.
SessionAgentDeselectParams: Identifies the target session.
SessionAgentGetCurrentParams: Identifies the target session.
SessionAgentGetCurrentResult: The currently selected custom agent, or null when using the default agent.
SessionAgentListParams: Identifies the target session.
SessionAgentListResult: Custom agents available to the session.
SessionAgentReloadParams: Identifies the target session.
SessionAgentReloadResult: Custom agents available to the session after reloading definitions from disk.
SessionAgentSelectResult: The newly selected custom agent.
SessionAuthStatus: Authentication status and account metadata for the session.
SessionBulkDeleteResult: Map of sessionId -> bytes freed by removing the session’s workspace directory.
SessionCanvasActionInvokeResult: Canvas action invocation result.
SessionCanvasListOpenParams: Identifies the target session.
SessionCanvasListOpenResult: Live open-canvas snapshot.
SessionCanvasListParams: Identifies the target session.
SessionCanvasListResult: Declared canvases available in this session.
SessionCanvasOpenResult: Open canvas instance snapshot.
SessionCommandsEnqueueResult: Indicates whether the command was accepted into the local execution queue.
SessionCommandsExecuteResult: Error message produced while executing the command, if any.
SessionCommandsHandlePendingCommandResult: Indicates whether the pending client-handled command was completed successfully.
SessionCommandsListResult: Slash commands available in the session, after applying any include/exclude filters.
SessionCommandsRespondToQueuedCommandResult: Indicates whether the queued-command response was matched to a pending request.
SessionCompletionItem: A single host-driven completion. Accepting an item replaces [rangeStart, rangeEnd) (UTF-16 code units) in the composer with insertText; when the range is absent, the active token around the cursor is replaced.
SessionCompletionsGetTriggerCharactersParams: Identifies the target session.
SessionCompletionsGetTriggerCharactersResult: Characters that, when typed in the composer, should trigger a completions.request. Empty when the session has no host-driven completions (e.g. local sessions, or a relay host that does not advertise completionTriggerCharacters).
SessionCompletionsRequestResult: Host-driven completion items for the current composer input. Empty when the host returns no items or does not support completions.
SessionContext: Pre-resolved working-directory context for session startup.
SessionContextAttribution: Per-source context-window attribution, or null if the session has not yet been initialized (no system prompt or tool metadata cached).
SessionContextAttributionCompactions: Successful compaction history for the session.
SessionContextAttributionEntriesItem
SessionContextInfo: Token breakdown for the current context window, or null if the session has not yet been initialized (no system prompt or tool metadata cached).
SessionDebugCollectLogsResult: Result of collecting a redacted debug bundle.
SessionEnrichMetadataResult: The enriched metadata records, with summary and context fields backfilled where available. Sessions confirmed empty and unnamed are omitted.
SessionEventLogReadResult: Batch of session events returned by a read, with cursor and continuation metadata.
SessionEventLogRegisterInterestResult: Opaque handle representing an event-type interest registration.
SessionEventLogReleaseInterestResult: Indicates whether the operation succeeded.
SessionEventLogTailParams: Identifies the target session.
SessionEventLogTailResult: Snapshot of the current tail cursor without returning any events. Use this when a consumer wants to subscribe to live events going forward without first paginating through the entire persisted history (which would happen if read were called without a cursor on a long-lived session).
SessionExtensionsListParams: Identifies the target session.
SessionExtensionsListResult: Extensions discovered for the session, with their current status.
SessionExtensionsReloadParams: Identifies the target session.
SessionFleetStartResult: Indicates whether fleet mode was successfully activated.
SessionFsAppendFileRequest: File path, content to append, and optional mode for the client-provided session filesystem.
SessionFsError: Describes a filesystem error.
SessionFsExistsRequest: Path to test for existence in the client-provided session filesystem.
SessionFsExistsResult: Indicates whether the requested path exists in the client-provided session filesystem.
SessionFsMkdirRequest: Directory path to create in the client-provided session filesystem, with options for recursive creation and POSIX mode.
SessionFsReadFileRequest: Path of the file to read from the client-provided session filesystem.
SessionFsReadFileResult: File content as a UTF-8 string, or a filesystem error if the read failed.
SessionFsReaddirRequest: Directory path whose entries should be listed from the client-provided session filesystem.
SessionFsReaddirResult: Names of entries in the requested directory, or a filesystem error if the read failed.
SessionFsReaddirWithTypesEntry: Directory entry returned by session filesystem readdirWithTypes, with name and entry type.
SessionFsReaddirWithTypesRequest: Directory path whose entries (with type information) should be listed from the client-provided session filesystem.
SessionFsReaddirWithTypesResult: Entries in the requested directory paired with file/directory type information, or a filesystem error if the read failed.
SessionFsRenameRequest: Source and destination paths for renaming or moving an entry in the client-provided session filesystem.
SessionFsRmRequest: Path to remove from the client-provided session filesystem, with options for recursive removal and force.
SessionFsSetProviderCapabilities: Optional capabilities declared by the provider
SessionFsSetProviderRequest: Initial working directory, session-state path layout, and path conventions used to register the calling SDK client as the session filesystem provider.
SessionFsSetProviderResult: Indicates whether the calling client was registered as the session filesystem provider.
SessionFsSqliteExistsParams: Identifies the target session.
SessionFsSqliteExistsResult: Indicates whether the per-session SQLite database already exists.
SessionFsSqliteQueryRequest: SQL query, query type, and optional bind parameters for executing a SQLite query against the per-session database.
SessionFsSqliteQueryResult: Query results including rows, columns, and rows affected, or a filesystem error if execution failed.
SessionFsStatRequest: Path whose metadata should be returned from the client-provided session filesystem.
SessionFsStatResult: Filesystem metadata for the requested path, or a filesystem error if the stat failed.
SessionFsWriteFileRequest: File path, content to write, and optional mode for the client-provided session filesystem.
SessionGitHubAuthGetStatusParams: Identifies the target session.
SessionGitHubAuthGetStatusResult: Authentication status and account metadata for the session.
SessionGitHubAuthSetCredentialsResult: Indicates whether the credential update succeeded.
SessionHistoryAbortManualCompactionParams: Identifies the target session.
SessionHistoryAbortManualCompactionResult: Indicates whether an in-progress manual compaction was aborted.
SessionHistoryCancelBackgroundCompactionParams: Identifies the target session.
SessionHistoryCancelBackgroundCompactionResult: Indicates whether an in-progress background compaction was cancelled.
SessionHistoryCompactResult: Compaction outcome with the number of tokens and messages removed, summary text, and the resulting context window breakdown.
SessionHistorySummarizeForHandoffParams: Identifies the target session.
SessionHistorySummarizeForHandoffResult: Markdown summary of the conversation context (empty when not available).
SessionHistoryTruncateResult: Number of events that were removed by the truncation.
SessionInstalledPlugin: Installed plugin record for a session, with marketplace, version, install time, enabled state, cache path, and source.
SessionInstalledPluginSourceGitHub: Source descriptor for a direct GitHub plugin install, with owner/repo, optional ref, and optional subpath.
SessionInstalledPluginSourceLocal: Source descriptor for a direct local plugin install, with a local filesystem path.
SessionInstalledPluginSourceUrl: Source descriptor for a direct URL plugin install, with URL, optional ref, and optional subpath.
SessionInstructionsGetSourcesParams: Identifies the target session.
SessionInstructionsGetSourcesResult: Instruction sources loaded for the session, in merge order.
SessionList: Sessions matching the filter, ordered most-recently-modified first.
SessionListFilter: Optional filter applied to the returned sessions
SessionLoadDeferredRepoHooksResult: Queued repo-level startup prompts and the total hook command count after loading.
SessionLogResult: Identifier of the session event that was emitted for the log message.
SessionMcpAppsDiagnoseResult: Diagnostic snapshot of MCP Apps wiring for the named server.
SessionMcpAppsGetHostContextParams: Identifies the target session.
SessionMcpAppsGetHostContextResult: Current host context advertised to MCP App guests.
SessionMcpAppsListToolsResult: App-callable tools from the named MCP server.
SessionMcpAppsReadResourceResult: Resource contents returned by the MCP server.
SessionMcpCancelSamplingExecutionResult: Indicates whether an in-flight sampling execution with the given requestId was found and cancelled.
SessionMcpConfigureGitHubResult: Result of configuring GitHub MCP.
SessionMcpExecuteSamplingResult: Outcome of an MCP sampling execution: success result, failure error, or cancellation.
SessionMcpHeadersHandlePendingHeadersRefreshRequestResult: Indicates whether the pending MCP headers refresh response was accepted.
SessionMcpIsServerRunningResult: Whether the named MCP server is running.
SessionMcpListParams: Identifies the target session.
SessionMcpListResult: MCP servers configured for the session, with their connection status and host-level state.
SessionMcpListToolsResult: Tools exposed by the connected MCP server. Throws when the server is not connected.
SessionMcpOauthHandlePendingRequestResult: Indicates whether the pending MCP OAuth response was accepted.
SessionMcpOauthLoginResult: OAuth authorization URL the caller should open, or empty when cached tokens already authenticated the server.
SessionMcpReloadParams: Identifies the target session.
SessionMcpReloadWithConfigResult: MCP server startup filtering result.
SessionMcpRemoveGitHubParams: Identifies the target session.
SessionMcpRemoveGitHubResult: Indicates whether the auto-managed github MCP server was removed (false when nothing to remove).
SessionMcpSetEnvValueModeResult: Env-value mode recorded on the session after the update.
SessionMetadataActivityParams: Identifies the target session.
SessionMetadataActivityResult: Current activity flags for the session.
SessionMetadataContextInfoResult: Token breakdown for the session’s current context window, or null if uninitialized.
SessionMetadataContextInfoResultContextInfo: Token-usage breakdown for the session’s current context window
SessionMetadataGetContextAttributionParams: Identifies the target session.
SessionMetadataGetContextAttributionResult: Per-source attribution breakdown for the session’s current context window, or null if uninitialized.
SessionMetadataGetContextAttributionResultContextAttribution: Per-source token attribution snapshot for the current context window. The heaviest individual messages are available separately via metadata.getContextHeaviestMessages.
SessionMetadataGetContextAttributionResultContextAttributionCompactions: Successful compaction history for the session.
SessionMetadataGetContextAttributionResultContextAttributionEntriesItem
SessionMetadataGetContextHeaviestMessagesResult: The heaviest individual messages in the session’s context window, most-expensive first.
SessionMetadataIsProcessingParams: Identifies the target session.
SessionMetadataIsProcessingResult: Indicates whether the local session is currently processing a turn or background continuation.
SessionMetadataRecomputeContextTokensResult: Re-tokenize the session’s existing messages against modelId and return the token totals. Useful for hosts that want an initial estimate of context usage on session resume, before the next agent turn fires session.context_info_changed events. Returns zeros for an empty session.
SessionMetadataRecordContextChangeResult: Notify the session that its working directory context has changed. Emits a session.context_changed event so consumers (telemetry, OTel tracker, ACP, the timeline UI) can react. Use this when the host has detected a cwd/branch/repo change outside the session’s normal lifecycle (e.g., after a shell command in interactive mode).
SessionMetadataSetWorkingDirectoryResult: Update the session’s working directory. Used by the host when the user explicitly changes cwd (e.g., the /cd slash command). The host is responsible for process.chdir and any related side-effects (file index, etc.); this method only updates the session’s own recorded path.
SessionMetadataSnapshot: Point-in-time snapshot of slow-changing session identifier and state fields
SessionMetadataSnapshotParams: Identifies the target session.
SessionMetadataSnapshotResult: Point-in-time snapshot of slow-changing session identifier and state fields
SessionMetadataSnapshotResultWorkspace: Public-facing projection of workspace metadata for SDK / TUI consumers
SessionMetadataSnapshotWorkspace: Public-facing projection of workspace metadata for SDK / TUI consumers
SessionModeGetParams: Identifies the target session.
SessionModelGetCurrentParams: Identifies the target session.
SessionModelGetCurrentResult: The currently selected model, reasoning effort, and context tier for the session. The context tier reflects Session.getContextTier(), restored from the session journal on resume.
SessionModelList: The list of models available to this session.
SessionModelListResult: The list of models available to this session.
SessionModelSetReasoningEffortResult: Update the session’s reasoning effort without changing the selected model. Use switchTo instead when you also need to change the model. The runtime stores the effort on the session and applies it to subsequent turns.
SessionModelSwitchToResult: The model identifier active on the session after the switch.
SessionNameGetParams: Identifies the target session.
SessionNameGetResult: The session’s friendly name, or null when not yet set.
SessionNameSetAutoResult: Indicates whether the auto-generated summary was applied as the session’s name.
SessionOpenOptions: Session construction options.
SessionOpenOptionsAdditionalContentExclusionPolicy: Content-exclusion policy supplied to sessions.open options, with rules, last-updated data, and scope.
SessionOpenOptionsAdditionalContentExclusionPolicyRule: Single content-exclusion rule supplied to sessions.open options, with paths, match conditions, and source.
SessionOpenOptionsAdditionalContentExclusionPolicyRuleSource: Source descriptor for a sessions.open content-exclusion rule, with source name and type.
SessionOpenResult: Result of opening a session.
SessionOptionsUpdateResult: Indicates whether the session options patch was applied successfully.
SessionPermissionsConfigureResult: Indicates whether the operation succeeded.
SessionPermissionsFolderTrustAddTrustedResult: Indicates whether the operation succeeded.
SessionPermissionsFolderTrustIsTrustedResult: Folder trust check result.
SessionPermissionsGetAllowAllResult: Current allow-all permission mode.
SessionPermissionsHandlePendingPermissionRequestResult: Indicates whether the permission decision was applied; false when the request was already resolved.
SessionPermissionsLocationsAddToolApprovalResult: Indicates whether the operation succeeded.
SessionPermissionsLocationsApplyResult: Summary of persisted location permissions applied to the session.
SessionPermissionsLocationsResolveResult: Resolved location-permissions key and type.
SessionPermissionsModifyRulesResult: Indicates whether the operation succeeded.
SessionPermissionsNotifyPromptShownResult: Indicates whether the operation succeeded.
SessionPermissionsPathsAddResult: Indicates whether the operation succeeded.
SessionPermissionsPathsIsPathWithinAllowedDirectoriesResult: Indicates whether the supplied path is within the session’s allowed directories.
SessionPermissionsPathsIsPathWithinWorkspaceResult: Indicates whether the supplied path is within the session’s workspace directory.
SessionPermissionsPathsListResult: Snapshot of the session’s allow-listed directories and primary working directory.
SessionPermissionsPathsUpdatePrimaryResult: Indicates whether the operation succeeded.
SessionPermissionsPendingRequestsResult: List of pending permission requests reconstructed from event history.
SessionPermissionsResetSessionApprovalsResult: Indicates whether the operation succeeded.
SessionPermissionsSetAllowAllResult: Indicates whether the operation succeeded and reports the post-mutation state.
SessionPermissionsSetApproveAllResult: Indicates whether the operation succeeded.
SessionPermissionsSetRequiredResult: Indicates whether the operation succeeded.
SessionPermissionsUrlsSetUnrestrictedModeResult: Indicates whether the operation succeeded.
SessionPlanDeleteParams: Identifies the target session.
SessionPlanReadParams: Identifies the target session.
SessionPlanReadResult: Existence, contents, and resolved path of the session plan file.
SessionPlanReadSqlTodosParams: Identifies the target session.
SessionPlanReadSqlTodosResult: Todo rows read from the session SQL database. Empty when no session database is available.
SessionPlanReadSqlTodosWithDependenciesParams: Identifies the target session.
SessionPlanReadSqlTodosWithDependenciesResult: Todo rows + dependency edges read from the session SQL database.
SessionPluginsListParams: Identifies the target session.
SessionPluginsListResult: Plugins installed for the session, with their enabled state and version metadata.
SessionProviderAddResult: The selectable model entries synthesized for the models added by this call.
SessionProviderGetEndpointResult: A snapshot of the provider endpoint the session is currently configured to talk to.
SessionPruneResult: Outcome of the prune operation: deleted IDs, dry-run candidates, skipped IDs, total bytes freed, and the dry-run flag.
SessionQueueClearParams: Identifies the target session.
SessionQueuePendingItemsParams: Identifies the target session.
SessionQueuePendingItemsResult: Snapshot of the session’s pending queued items and immediate-steering messages.
SessionQueueRemoveMostRecentParams: Identifies the target session.
SessionQueueRemoveMostRecentResult: Indicates whether a user-facing pending item was removed.
SessionRemoteDisableParams: Identifies the target session.
SessionRemoteEnableResult: GitHub URL for the session and a flag indicating whether remote steering is enabled.
SessionRemoteNotifySteerableChangedResult: Persist a steerability change as a session.remote_steerable_changed event. Used by the host (CLI / SDK consumer) when it has just finished enabling or disabling steering on a remote exporter that the runtime does not directly own.
SessionRpc: Typed view over a Session’s RPC namespace.
SessionRpcAgent: session.agent.* RPCs.
SessionRpcCanvas: session.canvas.* RPCs.
SessionRpcCanvasAction: session.canvas.action.* RPCs.
SessionRpcCommands: session.commands.* RPCs.
SessionRpcCompletions: session.completions.* RPCs.
SessionRpcDebug: session.debug.* RPCs.
SessionRpcEventLog: session.eventLog.* RPCs.
SessionRpcExtensions: session.extensions.* RPCs.
SessionRpcFleet: session.fleet.* RPCs.
SessionRpcGitHubAuth: session.gitHubAuth.* RPCs.
SessionRpcHistory: session.history.* RPCs.
SessionRpcInstructions: session.instructions.* RPCs.
SessionRpcLsp: session.lsp.* RPCs.
SessionRpcMcp: session.mcp.* RPCs.
SessionRpcMcpApps: session.mcp.apps.* RPCs.
SessionRpcMcpHeaders: session.mcp.headers.* RPCs.
SessionRpcMcpOauth: session.mcp.oauth.* RPCs.
SessionRpcMetadata: session.metadata.* RPCs.
SessionRpcMode: session.mode.* RPCs.
SessionRpcModel: session.model.* RPCs.
SessionRpcName: session.name.* RPCs.
SessionRpcOptions: session.options.* RPCs.
SessionRpcPermissions: session.permissions.* RPCs.
SessionRpcPermissionsFolderTrust: session.permissions.folderTrust.* RPCs.
SessionRpcPermissionsLocations: session.permissions.locations.* RPCs.
SessionRpcPermissionsPaths: session.permissions.paths.* RPCs.
SessionRpcPermissionsUrls: session.permissions.urls.* RPCs.
SessionRpcPlan: session.plan.* RPCs.
SessionRpcPlugins: session.plugins.* RPCs.
SessionRpcProvider: session.provider.* RPCs.
SessionRpcQueue: session.queue.* RPCs.
SessionRpcRemote: session.remote.* RPCs.
SessionRpcSchedule: session.schedule.* RPCs.
SessionRpcSettings: session.settings.* RPCs.
SessionRpcShell: session.shell.* RPCs.
SessionRpcSkills: session.skills.* RPCs.
SessionRpcTasks: session.tasks.* RPCs.
SessionRpcTelemetry: session.telemetry.* RPCs.
SessionRpcTools: session.tools.* RPCs.
SessionRpcUi: session.ui.* RPCs.
SessionRpcUsage: session.usage.* RPCs.
SessionRpcVisibility: session.visibility.* RPCs.
SessionRpcWorkspaces: session.workspaces.* RPCs.
SessionScheduleListParams: Identifies the target session.
SessionScheduleListResult: Snapshot of the currently active recurring prompts for this session.
SessionScheduleStopResult: Remove a scheduled prompt by id. The result entry is omitted if the id was unknown.
SessionSendResult: Result of sending a user message
SessionSetCredentialsParams: New auth credentials to install on the session. Omit to leave credentials unchanged.
SessionSetCredentialsResult: Indicates whether the credential update succeeded.
SessionSettingsBuiltInToolAvailabilitySnapshot: Availability of built-in job tools surfaced to boundary consumers.
SessionSettingsEvaluatePredicateRequest: Named Rust-owned settings predicate to evaluate for this session.
SessionSettingsEvaluatePredicateResult: Result of evaluating a Rust-owned settings predicate.
SessionSettingsJobSnapshot: Redacted job settings for a session. The job nonce is excluded.
SessionSettingsModelSnapshot: Redacted model routing settings for a session.
SessionSettingsOnlineEvaluationSnapshot: Online-evaluation settings safe to expose across the SDK boundary.
SessionSettingsRepoSnapshot: Redacted repository and GitHub host settings for a session.
SessionSettingsSnapshot: Redacted, serializable view of session runtime settings for SDK boundary consumers. Secrets and raw feature flags are intentionally excluded.
SessionSettingsSnapshotParams: Identifies the target session.
SessionSettingsSnapshotResult: Redacted, serializable view of session runtime settings for SDK boundary consumers. Secrets and raw feature flags are intentionally excluded.
SessionSettingsValidationSnapshot: Redacted validation and memory-tool settings for a session.
SessionShellCancelUserRequestedResult: Cancellation result for a user-requested shell command.
SessionShellExecResult: Identifier of the spawned process, used to correlate streamed output and exit notifications.
SessionShellExecuteUserRequestedResult: Result of a user-requested shell command.
SessionShellKillResult: Indicates whether the signal was delivered; false if the process was unknown or already exited.
SessionSizes: Map of sessionId -> on-disk size in bytes for each session’s workspace directory.
SessionSkillsEnsureLoadedParams: Identifies the target session.
SessionSkillsGetInvokedParams: Identifies the target session.
SessionSkillsGetInvokedResult: Skills invoked during this session, ordered by invocation time (most recent last).
SessionSkillsListParams: Identifies the target session.
SessionSkillsListResult: Skills available to the session, with their enabled state.
SessionSkillsReloadParams: Identifies the target session.
SessionSkillsReloadResult: Diagnostics from reloading skill definitions, with warnings and errors as separate lists.
SessionSuspendParams: Identifies the target session.
SessionTasksCancelResult: Indicates whether the background task was successfully cancelled.
SessionTasksGetCurrentPromotableParams: Identifies the target session.
SessionTasksGetCurrentPromotableResult: The first sync-waiting task that can currently be promoted to background mode.
SessionTasksGetProgressResult: Progress information for the task, or null when no task with that ID is tracked.
SessionTasksListParams: Identifies the target session.
SessionTasksListResult: Background tasks currently tracked by the session.
SessionTasksPromoteCurrentToBackgroundParams: Identifies the target session.
SessionTasksPromoteCurrentToBackgroundResult: The promoted task as it now exists in background mode, omitted if no promotable task was waiting.
SessionTasksPromoteToBackgroundResult: Indicates whether the task was successfully promoted to background mode.
SessionTasksRefreshParams: Identifies the target session.
SessionTasksRefreshResult: Refresh metadata for any detached background shells the runtime knows about. Use after a long pause to pick up exit/output state for shells running outside the agent loop.
SessionTasksRemoveResult: Indicates whether the task was removed. False when the task does not exist or is still running/idle.
SessionTasksSendMessageResult: Indicates whether the message was delivered, with an error message when delivery failed.
SessionTasksStartAgentResult: Identifier assigned to the newly started background agent task.
SessionTasksWaitForPendingParams: Identifies the target session.
SessionTasksWaitForPendingResult: Wait until all in-flight background tasks (agents + shells) and any follow-up turns scheduled by their completions have settled. Returns when the runtime is fully drained or after an internal timeout (default 10 minutes; configurable via COPILOT_TASK_WAIT_TIMEOUT_SECONDS).
SessionTelemetryEngagement: Telemetry engagement ID for the session, when available.
SessionTelemetryGetEngagementIdParams: Identifies the target session.
SessionTelemetryGetEngagementIdResult: Telemetry engagement ID for the session, when available.
SessionToolsGetCurrentMetadataParams: Identifies the target session.
SessionToolsGetCurrentMetadataResult: Current lightweight tool metadata snapshot for the session.
SessionToolsHandlePendingToolCallResult: Indicates whether the external tool call result was handled successfully.
SessionToolsInitializeAndValidateParams: Identifies the target session.
SessionToolsInitializeAndValidateResult: Resolve, build, and validate the runtime tool list for this session. Subagent sessions and consumer flows that need an initialized tool set before send invoke this. Default base-class implementation is a no-op for sessions that don’t support tool validation.
SessionToolsUpdateSubagentSettingsResult: Empty result after applying subagent settings
SessionUiElicitationResult: The elicitation response (accept with form values, decline, or cancel)
SessionUiEphemeralQueryResult: Transient answer generated from current conversation context.
SessionUiHandlePendingAutoModeSwitchResult: Indicates whether the pending UI request was resolved by this call.
SessionUiHandlePendingElicitationResult: Indicates whether the elicitation response was accepted; false if it was already resolved by another client.
SessionUiHandlePendingExitPlanModeResult: Indicates whether the pending UI request was resolved by this call.
SessionUiHandlePendingSamplingResult: Indicates whether the pending UI request was resolved by this call.
SessionUiHandlePendingSessionLimitsExhaustedResult: Indicates whether the pending UI request was resolved by this call.
SessionUiHandlePendingUserInputResult: Indicates whether the pending UI request was resolved by this call.
SessionUiRegisterDirectAutoModeSwitchHandlerParams: Identifies the target session.
SessionUiRegisterDirectAutoModeSwitchHandlerResult: Register an in-process handler for auto_mode_switch.requested events. The caller still attaches the actual listener via the standard event-subscription mechanism; this registration solely tells the server bridge to skip its own dispatch (so a remote client doesn’t race the in-process handler for the same requestId).
SessionUiUnregisterDirectAutoModeSwitchHandlerResult: Indicates whether the handle was active and the registration count was decremented.
SessionUpdateOptionsParams: Patch of mutable session options to apply to the running session.
SessionUpdateOptionsResult: Indicates whether the session options patch was applied successfully.
SessionUsageGetMetricsParams: Identifies the target session.
SessionUsageGetMetricsResult: Accumulated session usage metrics, including premium request cost, token counts, model breakdown, and code-change totals.
SessionVisibilityGetParams: Identifies the target session.
SessionVisibilityGetResult: Current sharing status and shareable GitHub URL for a session.
SessionVisibilitySetResult: Effective sharing status and shareable GitHub URL after updating session visibility.
SessionWorkingDirectoryContext: Updated working directory and git context. Emitted as the new payload of session.context_changed.
SessionWorkspacesDiffResult: Workspace diff result for the requested mode.
SessionWorkspacesGetWorkspaceParams: Identifies the target session.
SessionWorkspacesGetWorkspaceResult: Current workspace metadata for the session, including its absolute filesystem path when available.
SessionWorkspacesGetWorkspaceResultWorkspace
SessionWorkspacesListCheckpointsParams: Identifies the target session.
SessionWorkspacesListCheckpointsResult: Workspace checkpoints in chronological order; empty when the workspace is not enabled.
SessionWorkspacesListFilesParams: Identifies the target session.
SessionWorkspacesListFilesResult: Relative paths of files stored in the session workspace files directory.
SessionWorkspacesReadCheckpointResult: Checkpoint content as a UTF-8 string, or null when the checkpoint or workspace is missing.
SessionWorkspacesReadFileResult: Contents of the requested workspace file as a UTF-8 string.
SessionWorkspacesSaveLargePasteResult: Descriptor for the saved paste file, or null when the workspace is unavailable.
SessionWorkspacesSaveLargePasteResultSaved
SessionsBulkDeleteRequest: Session IDs to close, deactivate, and delete from disk.
SessionsBulkDeleteResult: Map of sessionId -> bytes freed by removing the session’s workspace directory.
SessionsCheckInUseRequest: Session IDs to test for live in-use locks.
SessionsCheckInUseResult: Session IDs from the input set that are currently in use by another process.
SessionsCloseRequest: Session ID to close.
SessionsCloseResult: Closes a session: emits shutdown, flushes pending events to disk, releases the in-use lock, disposes the active session. Idempotent: succeeds even if the session is not currently active.
SessionsConnectResult: Remote session connection result.
SessionsEnrichMetadataRequest: Session metadata records to enrich with summary and context information.
SessionsEnrichMetadataResult: The enriched metadata records, with summary and context fields backfilled where available. Sessions confirmed empty and unnamed are omitted.
SessionsFindByPrefixRequest: UUID prefix to resolve to a unique session ID.
SessionsFindByPrefixResult: Session ID matching the prefix, omitted when no unique match exists.
SessionsFindByTaskIDRequest: GitHub task ID to look up.
SessionsFindByTaskIDResult: ID of the local session bound to the given GitHub task, or omitted when none.
SessionsFindByTaskIdResult: ID of the local session bound to the given GitHub task, or omitted when none.
SessionsForkRequest: Source session identifier to fork from, optional event-ID boundary, and optional friendly name for the new session.
SessionsForkResult: Identifier and optional friendly name assigned to the newly forked session.
SessionsGetBoardEntryCountRequest: Session ID whose board entry count should be returned.
SessionsGetBoardEntryCountResult: Dynamic-context board entry count, when available.
SessionsGetEventFilePathRequest: Session ID whose event-log file path to compute.
SessionsGetEventFilePathResult: Absolute path to the session’s events.jsonl file on disk.
SessionsGetLastForContextRequest: Optional working-directory context used to score session relevance.
SessionsGetLastForContextResult: Most-relevant session ID for the supplied context, or omitted when no sessions exist.
SessionsGetPersistedRemoteSteerableRequest: Session ID to look up the persisted remote-steerable flag for.
SessionsGetPersistedRemoteSteerableResult: The session’s persisted remote-steerable flag, or omitted when no value has been persisted.
SessionsGetRemoteControlStatusResult: Wrapper for the singleton’s current status.
SessionsGetSizesResult: Map of sessionId -> on-disk size in bytes for each session’s workspace directory.
SessionsListRequest: Optional source filter, metadata-load limit, and context filter applied to the returned sessions.
SessionsListResult: Sessions matching the filter, ordered most-recently-modified first.
SessionsLoadDeferredRepoHooksRequest: Active session ID whose deferred repo-level hooks should be loaded.
SessionsLoadDeferredRepoHooksResult: Queued repo-level startup prompts and the total hook command count after loading.
SessionsOpenAttach: Parameters for attaching to an already-active session by ID.
SessionsOpenCloud: Parameters for creating a new cloud session.
SessionsOpenCreate: Parameters for creating a new local session.
SessionsOpenHandoff: Parameters for fetching a remote session and handing it off to a new local session.
SessionsOpenProgress: sessions.open handoff progress update with step, status, and optional message.
SessionsOpenRemote: Parameters for connecting to a live remote session.
SessionsOpenResult: Result of opening a session.
SessionsOpenResume: Parameters for resuming a specific local session.
SessionsOpenResumeLast: Parameters for resuming the most relevant local session.
SessionsPruneOldRequest: Age threshold and optional flags controlling which old sessions are pruned (or simulated when dryRun is true).
SessionsPruneOldResult: Outcome of the prune operation: deleted IDs, dry-run candidates, skipped IDs, total bytes freed, and the dry-run flag.
SessionsRegisterExtensionToolsOnSessionOptions: Optional registration options.
SessionsReleaseLockRequest: Session ID whose in-use lock should be released.
SessionsReleaseLockResult: Release the in-use lock held by this process for the given session. No-op when this process does not currently hold a lock for the session.
SessionsReloadPluginHooksRequest: Active session ID and an optional flag for deferring repo-level hooks until folder trust.
SessionsReloadPluginHooksResult: Reload all hooks (user, plugin, optionally repo) and apply them to the active session. Call after installing or removing plugins so their hooks take effect immediately. No-op when no active session matches the given sessionId.
SessionsSaveRequest: Session ID whose pending events should be flushed to disk.
SessionsSaveResult: Flush a session’s pending events to disk. No-op when no writer exists for the session (e.g., already closed).
SessionsSetAdditionalPluginsRequest: Manager-wide additional plugins to register; replaces any previously-configured set.
SessionsSetAdditionalPluginsResult: Replace the manager-wide additional plugins. New session creations and subsequent hook reloads see the new set; already-running sessions keep their existing hook installation until the next reload.
SessionsSetRemoteControlSteeringRequest: Patch for the singleton’s steering state.
SessionsSetRemoteControlSteeringResult: Wrapper for the singleton’s current status.
SessionsStartRemoteControlRequest: Parameters for attaching the remote-control singleton to a session.
SessionsStartRemoteControlResult: Wrapper for the singleton’s current status.
SessionsStopRemoteControlRequest: Parameters for stopping the remote-control singleton.
SessionsStopRemoteControlResult: Outcome of a stopRemoteControl call.
SessionsTransferRemoteControlRequest: Parameters for atomically rebinding the remote-control singleton.
SessionsTransferRemoteControlResult: Outcome of a transferRemoteControl call.
ShellCancelUserRequestedRequest: User-requested shell execution cancellation handle.
ShellExecRequest: Shell command to run, with optional working directory and timeout in milliseconds.
ShellExecResult: Identifier of the spawned process, used to correlate streamed output and exit notifications.
ShellExecuteUserRequestedRequest: User-requested shell command and cancellation handle.
ShellKillRequest: Identifier of a process previously returned by “shell.exec” and the signal to send.
ShellKillResult: Indicates whether the signal was delivered; false if the process was unknown or already exited.
ShutdownRequest: Parameters for shutting down the session
Skill: Skill metadata available to a session, with name, description, source, enabled/invocable state, path, plugin, and argument hint.
SkillDiscoveryPath: Canonical directory where skills can be discovered or created, with scope, preference, and optional project path.
SkillDiscoveryPathList: Canonical locations where skills can be created so the runtime will recognize them.
SkillList: Skills available to the session, with their enabled state.
SkillsConfigSetDisabledSkillsRequest: Skill names to mark as disabled in global configuration, replacing any previous list.
SkillsDisableRequest: Name of the skill to disable for the session.
SkillsDiscoverRequest: Optional project paths and additional skill directories to include in discovery.
SkillsDiscoverResult: Skills discovered across global and project sources.
SkillsEnableRequest: Name of the skill to enable for the session.
SkillsGetDiscoveryPathsRequest: Optional project paths to enumerate.
SkillsGetDiscoveryPathsResult: Canonical locations where skills can be created so the runtime will recognize them.
SkillsGetInvokedResult: Skills invoked during this session, ordered by invocation time (most recent last).
SkillsInvokedSkill: Skill invocation record with name, path, content, allowed tools, and turn number.
SkillsLoadDiagnostics: Diagnostics from reloading skill definitions, with warnings and errors as separate lists.
SlashCommandAgentPromptResult: Slash-command invocation result that submits an agent prompt, with display prompt, optional mode, and settings-change flag.
SlashCommandCompletedResult: Slash-command invocation result indicating completion, with optional message and settings-change flag.
SlashCommandInfo: Slash-command metadata with name, aliases, description, kind, input hint, execution allowance, and schedulability.
SlashCommandInput: Optional unstructured input hint
SlashCommandInputChoice: A literal choice the command input accepts, with a human-facing description
SlashCommandSelectSubcommandOption: Selectable slash-command subcommand option with name, description, and optional group label.
SlashCommandSelectSubcommandResult: Slash-command invocation result asking the client to present subcommand options for a parent command.
SlashCommandTextResult: Slash-command invocation result containing text output plus Markdown/ANSI rendering flags.
SubagentSettings: Subagent settings to apply, or null to clear the live session override
SubagentSettingsEntry: Subagent model, reasoning effort, and context tier settings
TaskAgentInfo: Tracked background agent task metadata, including IDs, status, timing, agent type, prompt, model, result, and latest response.
TaskAgentProgress: Progress snapshot for an agent task, with recent activity lines and optional latest intent.
TaskList: Background tasks currently tracked by the session.
TaskProgressLine: Timestamped display line for task progress output or recent agent activity.
TaskShellInfo: Tracked shell task metadata, including ID, command, status, timing, attachment/execution mode, log path, and PID.
TaskShellProgress: Progress snapshot for a shell task, with recent stdout/stderr output and optional process ID.
TasksCancelRequest: Identifier of the background task to cancel.
TasksCancelResult: Indicates whether the background task was successfully cancelled.
TasksGetCurrentPromotableResult: The first sync-waiting task that can currently be promoted to background mode.
TasksGetProgressRequest: Identifier of the background task to fetch progress for.
TasksGetProgressResult: Progress information for the task, or null when no task with that ID is tracked.
TasksPromoteCurrentToBackgroundResult: The promoted task as it now exists in background mode, omitted if no promotable task was waiting.
TasksPromoteToBackgroundRequest: Identifier of the task to promote to background mode.
TasksPromoteToBackgroundResult: Indicates whether the task was successfully promoted to background mode.
TasksRefreshResult: Refresh metadata for any detached background shells the runtime knows about. Use after a long pause to pick up exit/output state for shells running outside the agent loop.
TasksRemoveRequest: Identifier of the completed or cancelled task to remove from tracking.
TasksRemoveResult: Indicates whether the task was removed. False when the task does not exist or is still running/idle.
TasksSendMessageRequest: Identifier of the target agent task, message content, and optional sender agent ID.
TasksSendMessageResult: Indicates whether the message was delivered, with an error message when delivery failed.
TasksStartAgentRequest: Agent type, prompt, name, and optional description and model override for the new task.
TasksStartAgentResult: Identifier assigned to the newly started background agent task.
TasksWaitForPendingResult: Wait until all in-flight background tasks (agents + shells) and any follow-up turns scheduled by their completions have settled. Returns when the runtime is fully drained or after an internal timeout (default 10 minutes; configurable via COPILOT_TASK_WAIT_TIMEOUT_SECONDS).
TelemetrySetFeatureOverridesRequest: Feature override key/value pairs to attach to subsequent telemetry events from this session.
TokenAuthInfo: Authentication-info variant for SDK-configured token authentication, carrying host and the secret token value.
Tool: Built-in tool metadata with identifier, optional namespaced name, description, input-parameter schema, and usage instructions.
ToolList: Built-in tools available for the requested model, with their parameters and instructions.
ToolsGetCurrentMetadataResult: Current lightweight tool metadata snapshot for the session.
ToolsInitializeAndValidateResult: Resolve, build, and validate the runtime tool list for this session. Subagent sessions and consumer flows that need an initialized tool set before send invoke this. Default base-class implementation is a no-op for sessions that don’t support tool validation.
ToolsListRequest: Optional model identifier whose tool overrides should be applied to the listing.
ToolsListResult: Built-in tools available for the requested model, with their parameters and instructions.
ToolsUpdateSubagentSettingsResult: Empty result after applying subagent settings
UIElicitationArrayAnyOfField: Multi-select string field where each option pairs a value with a display label.
UIElicitationArrayAnyOfFieldItems: Schema applied to each item in the array.
UIElicitationArrayAnyOfFieldItemsAnyOf: Selectable option for a UI elicitation multi-select array item, with submitted value and display label.
UIElicitationArrayEnumField: Multi-select string field whose allowed values are defined inline.
UIElicitationArrayEnumFieldItems: Schema applied to each item in the array.
UIElicitationRequest: Prompt message and JSON schema describing the form fields to elicit from the user.
UIElicitationResponse: The elicitation response (accept with form values, decline, or cancel)
UIElicitationResult: Indicates whether the elicitation response was accepted; false if it was already resolved by another client.
UIElicitationSchema: JSON Schema describing the form fields to present to the user
UIElicitationSchemaPropertyBoolean: Boolean field rendered as a yes/no toggle.
UIElicitationSchemaPropertyNumber: Numeric field accepting either a number or an integer.
UIElicitationSchemaPropertyString: Free-text string field with optional length and format constraints.
UIElicitationStringEnumField: Single-select string field whose allowed values are defined inline.
UIElicitationStringOneOfField: Single-select string field where each option pairs a value with a display label.
UIElicitationStringOneOfFieldOneOf: Selectable option for a UI elicitation single-select string field, with submitted value and display label.
UIEphemeralQueryRequest: Transient question to answer without adding it to conversation history.
UIEphemeralQueryResult: Transient answer generated from current conversation context.
UIExitPlanModeResponse: User response for a pending exit-plan-mode request, with approval state, selected action, auto-approve flag, and feedback.
UIHandlePendingAutoModeSwitchRequest: Request ID of a pending auto_mode_switch.requested event and the user’s response.
UIHandlePendingElicitationRequest: Pending elicitation request ID and the user’s response (accept/decline/cancel + form values).
UIHandlePendingExitPlanModeRequest: Request ID of a pending exit_plan_mode.requested event and the user’s response.
UIHandlePendingResult: Indicates whether the pending UI request was resolved by this call.
UIHandlePendingSamplingRequest: Request ID of a pending sampling.requested event and an optional sampling result payload (omit to reject).
UIHandlePendingSamplingResponse: Optional sampling result payload. Omit to reject/cancel the sampling request without providing a result.
UIHandlePendingSessionLimitsExhaustedRequest: Request ID of a pending session_limits_exhausted.requested event and the user’s selected limit action.
UIHandlePendingUserInputRequest: Request ID of a pending user_input.requested event and the user’s response.
UIRegisterDirectAutoModeSwitchHandlerResult: Register an in-process handler for auto_mode_switch.requested events. The caller still attaches the actual listener via the standard event-subscription mechanism; this registration solely tells the server bridge to skip its own dispatch (so a remote client doesn’t race the in-process handler for the same requestId).
UISessionLimitsExhaustedResponse: The user’s selected action for an exhausted session limit.
UIUnregisterDirectAutoModeSwitchHandlerRequest: Opaque handle previously returned by registerDirectAutoModeSwitchHandler to release.
UIUnregisterDirectAutoModeSwitchHandlerResult: Indicates whether the handle was active and the registration count was decremented.
UIUserInputResponse: User response for a pending user-input request, with answer text and whether it was typed freeform.
UpdateSubagentSettingsRequest: Subagent settings to apply to the current session
UpdateSubagentSettingsRequestSubagents: Configured per-agent subagent overrides
UsageGetMetricsResult: Accumulated session usage metrics, including premium request cost, token counts, model breakdown, and code-change totals.
UsageMetricsCodeChanges: Aggregated code change metrics
UsageMetricsModelMetric: Per-model usage metrics, including request counts/costs, token usage, nano-AI units, and per-token-type details.
UsageMetricsModelMetricRequests: Request count and cost metrics for this model
UsageMetricsModelMetricTokenDetail: Per-model token-detail entry containing the accumulated token count for one token type.
UsageMetricsModelMetricUsage: Token usage metrics for this model
UsageMetricsTokenDetail: Session-wide token-detail entry containing the accumulated token count for one token type.
UserAuthInfo: Authentication-info variant for OAuth user auth, with host and login; the token remains in the runtime secret store.
UserRequestedShellCommandResult: Result of a user-requested shell command.
UserSettingMetadata: A single user setting’s effective value alongside its default, so consumers can render settings left at their default.
UserSettingsGetResult: Per-key metadata for every known user setting (settings.json overlaid with the legacy config.json, config.json wins), including settings left at their default. Excludes repository- and enterprise-managed overrides.
UserSettingsSetRequest: Partial user settings to write to settings.json. Each top-level key is written individually, replacing the existing value; a key whose value is null is removed.
UserSettingsSetResult: Outcome of writing user settings.
VisibilityGetResult: Current sharing status and shareable GitHub URL for a session.
VisibilitySetRequest: Desired sharing status for the session.
VisibilitySetResult: Effective sharing status and shareable GitHub URL after updating session visibility.
WorkspaceDiffFileChange: A single changed file and its unified diff.
WorkspaceDiffResult: Workspace diff result for the requested mode.
WorkspaceSummary: Public-facing workspace metadata for this session, or null if the session has no associated workspace. Excludes runtime-internal fields (GitHub IDs, summary count, internal flags).
WorkspacesCheckpoints: Workspace checkpoint metadata with assigned number, human-readable title, and checkpoint filename.
WorkspacesCreateFileRequest: Relative path and UTF-8 content for the workspace file to create or overwrite.
WorkspacesDiffRequest: Parameters for computing a workspace diff.
WorkspacesGetWorkspaceResult: Current workspace metadata for the session, including its absolute filesystem path when available.
WorkspacesGetWorkspaceResultWorkspace
WorkspacesListCheckpointsResult: Workspace checkpoints in chronological order; empty when the workspace is not enabled.
WorkspacesListFilesResult: Relative paths of files stored in the session workspace files directory.
WorkspacesReadCheckpointRequest: Checkpoint number to read.
WorkspacesReadCheckpointResult: Checkpoint content as a UTF-8 string, or null when the checkpoint or workspace is missing.
WorkspacesReadFileRequest: Relative path of the workspace file to read.
WorkspacesReadFileResult: Contents of the requested workspace file as a UTF-8 string.
WorkspacesSaveLargePasteRequest: Pasted content to save as a UTF-8 file in the session workspace.
WorkspacesSaveLargePasteResult: Descriptor for the saved paste file, or null when the workspace is unavailable.
WorkspacesSaveLargePasteResultSaved

Enums§

AdaptiveThinkingSupport: Resolved Anthropic adaptive-thinking capability for a model.
AgentDiscoveryPathScope: Which tier this directory belongs to
AgentInfoSource: Where the agent definition was loaded from
AgentRegistryLiveTargetEntryAttentionKind: Kind of attention required when status === “attention”. Meaningful only when status === “attention”.
AgentRegistryLiveTargetEntryKind: Process kind tag for the registry entry
AgentRegistryLiveTargetEntryLastTerminalEvent: How the most recent turn ended (clean vs aborted). Lets the renderer distinguish done from done_cancelled.
AgentRegistryLiveTargetEntryStatus: Coarse lifecycle status of the foreground session
AgentRegistryLogCaptureOpenErrorReason: Categorized reason for log-open failure
AgentRegistrySpawnErrorKind: Discriminator: child_process.spawn itself failed
AgentRegistrySpawnPermissionMode: Permission posture for the new session. ‘yolo’ requires the controller-local session to currently be in allow-all mode.
AgentRegistrySpawnRegistryTimeoutKind: Discriminator: spawn succeeded but child never registered
AgentRegistrySpawnResult: Outcome of an agentRegistry.spawn call.
AgentRegistrySpawnSpawnedKind: Discriminator: managed-server child spawned successfully
AgentRegistrySpawnValidationErrorField: Which parameter field was invalid. Omitted when the rejection is not field-specific.
AgentRegistrySpawnValidationErrorKind: Discriminator: synchronous pre-validation rejected the request
AgentRegistrySpawnValidationErrorReason: Categorized reason for the rejection. Low-cardinality enum so telemetry can aggregate by reason without leaking raw paths or agent/model names.
ApiKeyAuthInfoType: API-key authentication for non-GitHub LLM providers (e.g. when running BYOM-style).
AttachmentBlobType: Attachment type discriminator
AttachmentDirectoryType: Attachment type discriminator
AttachmentExtensionContextType: Attachment type discriminator
AttachmentFileType: Attachment type discriminator
AttachmentGitHubActionsJobType: Attachment type discriminator
AttachmentGitHubCommitType: Attachment type discriminator
AttachmentGitHubFileDiffType: Attachment type discriminator
AttachmentGitHubFileType: Attachment type discriminator
AttachmentGitHubReferenceType: Type of GitHub reference
AttachmentGitHubReleaseType: Attachment type discriminator
AttachmentGitHubRepositoryType: Attachment type discriminator
AttachmentGitHubSnippetType: Attachment type discriminator
AttachmentGitHubTreeComparisonType: Attachment type discriminator
AttachmentGitHubUrlType: Attachment type discriminator
AttachmentSelectionType: Attachment type discriminator
AuthInfoType: Authentication type
ConnectedRemoteSessionMetadataKind: Neutral SDK discriminator for the connected remote session kind.
ContentFilterMode: Controls how MCP tool result content is filtered: none leaves content unchanged, markdown sanitizes HTML while preserving Markdown-friendly output, and hidden_characters removes characters that can hide directives.
CopilotApiTokenAuthInfoHost: Authentication host (always the public GitHub host).
CopilotApiTokenAuthInfoType: Direct Copilot API authentication via the GITHUB_COPILOT_API_TOKEN + COPILOT_API_URL environment-variable pair. The token itself is read from the environment by the runtime, not carried in this struct.
DebugCollectLogsDestination: Destination for the redacted debug bundle.
DebugCollectLogsDestinationArchiveKind
DebugCollectLogsDestinationDirectoryKind
DebugCollectLogsEntryKind: Kind of caller-provided debug log entry.
DebugCollectLogsRedaction: How a collected debug entry should be redacted before being staged.
DebugCollectLogsResultKind: Destination kind that was written.
DebugCollectLogsSource: Source category for a collected debug bundle entry.
DiscoveredMcpServerType: Server transport type: stdio, http, sse (deprecated), or memory
EnvAuthInfoType: Personal access token (PAT) or server-to-server token sourced from an environment variable.
EventsAgentScope: Agent-scope filter: ‘primary’ returns only main-agent events plus events whose type starts with ‘subagent.’ (matching the typed-subscription default behavior); ‘all’ returns events from all agents (matching wildcard-subscription behavior). Default is ‘all’ to preserve wildcard semantics for catch-up callers.
EventsCursorStatus: Cursor status: ‘ok’ means the cursor was applied successfully; ‘expired’ means the cursor referred to an event that no longer exists in history (e.g. truncated or compacted away) and the read started from the beginning of the remaining history.
ExtensionContextPushInputType: Attachment type discriminator
ExtensionSource: Discovery source: project (.github/extensions/), user (~/.copilot/extensions/), plugin (installed plugin), or session (session-state//extensions/)
ExtensionStatus: Current status: running, disabled, failed, or starting
ExternalToolTextResultForLlmBinaryResultsForLlmType: Binary result type discriminator. Use “image” for images and “resource” for other binary data.
ExternalToolTextResultForLlmContentAudioType: Content block type discriminator
ExternalToolTextResultForLlmContentImageType: Content block type discriminator
ExternalToolTextResultForLlmContentResourceLinkIconTheme: Theme variant this icon is intended for
ExternalToolTextResultForLlmContentResourceLinkType: Content block type discriminator
ExternalToolTextResultForLlmContentResourceType: Content block type discriminator
ExternalToolTextResultForLlmContentShellExitType: Content block type discriminator
ExternalToolTextResultForLlmContentTerminalType: Content block type discriminator
ExternalToolTextResultForLlmContentTextType: Content block type discriminator
GhCliAuthInfoType: Authentication via the gh CLI’s saved credentials.
HMACAuthInfoHost: Authentication host. HMAC auth always targets the public GitHub host.
HMACAuthInfoType: HMAC-based authentication used by GitHub-internal services.
InstalledPluginSourceGitHubSource: Constant value. Always “github”.
InstalledPluginSourceLocalSource: Constant value. Always “local”.
InstalledPluginSourceUrlSource: Constant value. Always “url”.
InstructionDiscoveryPathKind: Whether the target is a single file or a directory of instruction files
InstructionDiscoveryPathLocation: Which tier this target belongs to
InstructionSourceLocation: Where this source lives — used for UI grouping
InstructionSourceType: Category of instruction source — used for merge logic
LlmInferenceHttpRequestStartTransport: Transport the runtime would otherwise use for this request. http (the default when absent) covers plain HTTP and SSE responses; websocket indicates a full-duplex message channel where each body chunk maps to one WebSocket message and the binary flag distinguishes text from binary frames. The SDK consumer uses this to decide whether to service the request with an HTTP client or a WebSocket client. It is the one piece of request metadata the consumer cannot reliably infer from the URL or headers alone.
McpAppsHostContextDetailsAvailableDisplayMode: Allowed values for the McpAppsHostContextDetailsAvailableDisplayMode enumeration.
McpAppsHostContextDetailsDisplayMode: Current display mode (SEP-1865)
McpAppsHostContextDetailsPlatform: Platform type for responsive design
McpAppsHostContextDetailsTheme: UI theme preference per SEP-1865
McpAppsSetHostContextDetailsAvailableDisplayMode: Allowed values for the McpAppsSetHostContextDetailsAvailableDisplayMode enumeration.
McpAppsSetHostContextDetailsDisplayMode: Current display mode (SEP-1865)
McpAppsSetHostContextDetailsPlatform: Platform type for responsive design
McpAppsSetHostContextDetailsTheme: UI theme preference per SEP-1865
McpHeadersHandlePendingHeadersRefreshRequest: Host response: supply dynamic headers or decline this refresh.
McpHeadersHandlePendingHeadersRefreshRequestHeadersKind
McpHeadersHandlePendingHeadersRefreshRequestNoneKind
McpOauthLoginGrantType: OAuth grant type override for this login.
McpOauthPendingRequestResponse: Host response to the pending OAuth request.
McpOauthPendingRequestResponseCancelledKind
McpOauthPendingRequestResponseTokenKind
McpSamplingExecutionAction: Outcome of the sampling inference. ‘success’ produced a response; ‘failure’ encountered an error (including agent-side rejection by content filter or criteria); ‘cancelled’ the caller cancelled this execution via cancelSamplingExecution.
McpServerConfigDeferTools: Controls if tools provided by this server can be loaded on demand via tool search (auto) or always included in the initial tool list (never)
McpServerConfigHttpOauthGrantType: OAuth grant type to use when authenticating to the remote MCP server.
McpServerConfigHttpType: Remote transport type. Defaults to “http” when omitted.
McpSetEnvValueModeDetails: How environment-variable values supplied to MCP servers are resolved. “direct” passes literal string values; “indirect” treats values as references (e.g. names of environment variables on the host) that the runtime resolves before launch. Defaults to the runtime’s startup mode; clients that intentionally launch MCP servers with literal values (e.g. CLI prompt mode and ACP) set this to “direct”.
MetadataSnapshotCurrentMode: The current agent mode for this session (e.g., ‘interactive’, ‘plan’, ‘autopilot’)
MetadataSnapshotRemoteMetadataTaskType: Whether the remote task originated from Copilot Coding Agent (cca) or a CLI --remote invocation.
ModelPickerCategory: Model capability category for grouping in the model picker
ModelPickerPriceCategory: Relative cost tier for token-based billing users
ModelPolicyState: Current policy state for this model
OmittedBinaryOmittedReason: Why the binary data is absent: it exceeded the inline size limit, or its asset was unavailable
OptionsUpdateAdditionalContentExclusionPolicyScope: Allowed values for the OptionsUpdateAdditionalContentExclusionPolicyScope enumeration.
OptionsUpdateContextTier: Context tier for models with tiered pricing. The session uses this to derive effective modelCapabilitiesOverrides so compaction, truncation, token display, and request limits honor the selected tier.
OptionsUpdateEnvValueMode: How env values are passed to MCP servers (direct inlines literal values; indirect resolves at launch).
OptionsUpdateReasoningSummary: Reasoning summary mode for supported model clients.
OptionsUpdateToolFilterPrecedence: Controls how availableTools (allowlist) and excludedTools (denylist) combine when both are set.
PermissionDecision: The client’s response to the pending permission prompt
PermissionDecisionApproveForLocationApproval: Approval to persist for this location
PermissionDecisionApproveForLocationApprovalCommandsKind: Approval scoped to specific command identifiers.
PermissionDecisionApproveForLocationApprovalCustomToolKind: Approval covering a custom tool.
PermissionDecisionApproveForLocationApprovalExtensionManagementKind: Approval covering extension lifecycle operations such as enable, disable, or reload.
PermissionDecisionApproveForLocationApprovalExtensionPermissionAccessKind: Approval covering an extension’s request to access a permission-gated capability.
PermissionDecisionApproveForLocationApprovalMcpKind: Approval covering an MCP tool.
PermissionDecisionApproveForLocationApprovalMcpSamplingKind: Approval covering MCP sampling requests for a server.
PermissionDecisionApproveForLocationApprovalMemoryKind: Approval covering writes to long-term memory.
PermissionDecisionApproveForLocationApprovalReadKind: Approval covering read-only filesystem operations.
PermissionDecisionApproveForLocationApprovalWriteKind: Approval covering filesystem write operations.
PermissionDecisionApproveForLocationKind: Approve and persist for this project location
PermissionDecisionApproveForSessionApproval: Session-scoped approval to remember (tool prompts only; omitted for path/url prompts)
PermissionDecisionApproveForSessionApprovalCommandsKind: Approval scoped to specific command identifiers.
PermissionDecisionApproveForSessionApprovalCustomToolKind: Approval covering a custom tool.
PermissionDecisionApproveForSessionApprovalExtensionManagementKind: Approval covering extension lifecycle operations such as enable, disable, or reload.
PermissionDecisionApproveForSessionApprovalExtensionPermissionAccessKind: Approval covering an extension’s request to access a permission-gated capability.
PermissionDecisionApproveForSessionApprovalMcpKind: Approval covering an MCP tool.
PermissionDecisionApproveForSessionApprovalMcpSamplingKind: Approval covering MCP sampling requests for a server.
PermissionDecisionApproveForSessionApprovalMemoryKind: Approval covering writes to long-term memory.
PermissionDecisionApproveForSessionApprovalReadKind: Approval covering read-only filesystem operations.
PermissionDecisionApproveForSessionApprovalWriteKind: Approval covering filesystem write operations.
PermissionDecisionApproveForSessionKind: Approve and remember for the rest of the session
PermissionDecisionApproveOnceKind: Approve this single request only
PermissionDecisionApprovePermanentlyKind: Approve and persist across sessions (URL prompts only)
PermissionDecisionApprovedForLocationKind: Approved and persisted for this project location
PermissionDecisionApprovedForSessionKind: Approved and remembered for the rest of the session
PermissionDecisionApprovedKind: The permission request was approved
PermissionDecisionCancelledKind: The permission request was cancelled before a response was used
PermissionDecisionDeniedByContentExclusionPolicyKind: Denied by the organization’s content exclusion policy
PermissionDecisionDeniedByPermissionRequestHookKind: Denied by a permission request hook registered by an extension or plugin
PermissionDecisionDeniedByRulesKind: Denied because approval rules explicitly blocked it
PermissionDecisionDeniedInteractivelyByUserKind: Denied by the user during an interactive prompt
PermissionDecisionDeniedNoApprovalRuleAndCouldNotRequestFromUserKind: Denied because no approval rule matched and user confirmation was unavailable
PermissionDecisionRejectKind: Reject the request
PermissionDecisionUserNotAvailableKind: No user is available to confirm the request
PermissionLocationType: Whether the location is a git repo or directory
PermissionsAllowAllMode: Current or requested allow-all mode.
PermissionsConfigureAdditionalContentExclusionPolicyScope: Allowed values for the PermissionsConfigureAdditionalContentExclusionPolicyScope enumeration.
PermissionsLocationsAddToolApprovalDetails: Tool approval to persist and apply
PermissionsLocationsAddToolApprovalDetailsCommandsKind: Approval scoped to specific command identifiers.
PermissionsLocationsAddToolApprovalDetailsCustomToolKind: Approval covering a custom tool.
PermissionsLocationsAddToolApprovalDetailsExtensionManagementKind: Approval covering extension lifecycle operations such as enable, disable, or reload.
PermissionsLocationsAddToolApprovalDetailsExtensionPermissionAccessKind: Approval covering an extension’s request to access a permission-gated capability.
PermissionsLocationsAddToolApprovalDetailsMcpKind: Approval covering an MCP tool.
PermissionsLocationsAddToolApprovalDetailsMcpSamplingKind: Approval covering MCP sampling requests for a server.
PermissionsLocationsAddToolApprovalDetailsMemoryKind: Approval covering writes to long-term memory.
PermissionsLocationsAddToolApprovalDetailsReadKind: Approval covering read-only filesystem operations.
PermissionsLocationsAddToolApprovalDetailsWriteKind: Approval covering filesystem write operations.
PermissionsModifyRulesScope: Whether the change applies to ephemeral session-scoped rules (cleared at session end) or to location-scoped rules persisted via the location-permissions config file.
PermissionsSetAllowAllSource: Optional source for allow-all telemetry. Defaults to rpc when omitted for SDK callers.
PermissionsSetApproveAllSource: Optional source for allow-all telemetry. Defaults to rpc when omitted for SDK callers.
ProviderConfigTransport: Provider transport. Defaults to “http”.
ProviderConfigType: Provider type. Defaults to “openai” for generic OpenAI-compatible APIs.
ProviderConfigWireApi: Wire API format (openai/azure only). Defaults to “completions”.
ProviderEndpointTransport: Transport to be used for provider requests.
ProviderEndpointType: Provider family. Matches the type field of a BYOK provider config.
ProviderEndpointWireApi: Wire API to be used, when required for the provider type.
PushAttachmentBlobType: Attachment type discriminator
PushAttachmentDirectoryType: Attachment type discriminator
PushAttachmentFileType: Attachment type discriminator
PushAttachmentGitHubActionsJobType: Attachment type discriminator
PushAttachmentGitHubCommitType: Attachment type discriminator
PushAttachmentGitHubFileDiffType: Attachment type discriminator
PushAttachmentGitHubFileType: Attachment type discriminator
PushAttachmentGitHubReferenceType: Type of GitHub reference
PushAttachmentGitHubReleaseType: Attachment type discriminator
PushAttachmentGitHubRepositoryType: Attachment type discriminator
PushAttachmentGitHubSnippetType: Attachment type discriminator
PushAttachmentGitHubTreeComparisonType: Attachment type discriminator
PushAttachmentGitHubUrlType: Attachment type discriminator
PushAttachmentSelectionType: Attachment type discriminator
QueuePendingItemsKind: Whether this item is a queued user message or a queued slash command / model change
RemoteControlStatusActiveState: Remote control state tag: active.
RemoteControlStatusConnectingState: Remote control state tag: connecting.
RemoteControlStatusErrorState: Remote control state tag: setup failed.
RemoteControlStatusOffState: Remote control state tag: not connected.
RemoteSessionMetadataTaskType: Whether the remote task originated from CCA or CLI --remote.
RemoteSessionMode: Per-session remote mode. “off” disables remote, “export” exports session events to GitHub without enabling remote steering, “on” enables both export and remote steering.
SendAgentMode: The UI mode the agent was in when this message was sent. Defaults to the session’s current mode.
SendMode: How to deliver the message. enqueue (default) appends to the message queue. immediate interjects during an in-progress turn.
SessionCapability: Session capability enabled for this session
SessionContextHostType: Repository host type
SessionFsErrorCode: Error classification
SessionFsReaddirWithTypesEntryType: Entry type
SessionFsSetProviderConventions: Path conventions used by this filesystem
SessionFsSqliteQueryType: How to execute the query: ‘exec’ for DDL/multi-statement (no results), ‘query’ for SELECT (returns rows), ‘run’ for INSERT/UPDATE/DELETE (returns rowsAffected)
SessionInstalledPluginSourceGitHubSource: Constant value. Always “github”.
SessionInstalledPluginSourceLocalSource: Constant value. Always “local”.
SessionInstalledPluginSourceUrlSource: Constant value. Always “url”.
SessionLogLevel: Log severity level. Determines how the message is displayed in the timeline. Defaults to “info”.
SessionOpenOptionsAdditionalContentExclusionPolicyScope: Allowed values for the SessionOpenOptionsAdditionalContentExclusionPolicyScope enumeration.
SessionOpenOptionsEnvValueMode: How MCP server environment values are interpreted.
SessionOpenOptionsReasoningSummary: Initial reasoning summary mode for supported model clients.
SessionOpenParams: Open a session by creating, resuming, attaching, connecting to a remote, or handing off.
SessionSettingsPredicateName: Rust-owned settings predicates exposed across the SDK boundary. Raw feature-flag names are intentionally not part of the contract.
SessionSource: Which session sources to include. Defaults to local for backward compatibility.
SessionVisibilityStatus: Sharing status for a synced session. “repo” makes the session visible to anyone with read access to the repository; “unshared” restricts it to the creator and collaborators.
SessionWorkingDirectoryContextHostType: Hosting platform type of the repository
SessionsOpenAttachKind: Attach to an already-active in-process session by ID. Unlike resume, this does NOT re-load from disk; the session must already be loaded by an earlier create/resume call. Returns status: 'not_found' when no active session matches the id. Useful for in-process consumers that need a fresh API handle to a session opened elsewhere (e.g., a peer foreground-session switch).
SessionsOpenCloudKind: Create a new cloud (coding-agent) session.
SessionsOpenCreateKind: Create a new local session.
SessionsOpenHandoffKind: Fetch a remote session and hand it off to a new local session.
SessionsOpenHandoffTaskType: Task type determines the handoff strategy (CCA fetches events; CLI prepares a transient session).
SessionsOpenProgressStatus: Step status.
SessionsOpenProgressStep: Handoff step.
SessionsOpenRemoteKind: Connect to a live remote session.
SessionsOpenResumeKind: Resume a specific local session by ID or prefix.
SessionsOpenResumeLastKind: Resume the most relevant existing local session.
SessionsOpenStatus: Outcome of the open request.
ShellKillSignal: Signal to send (default: SIGTERM)
SkillDiscoveryScope: Which tier this directory belongs to
SlashCommandAgentPromptResultKind: Agent prompt result discriminator
SlashCommandCompletedResultKind: Completed result discriminator
SlashCommandInputCompletion: Optional completion hint for the input (e.g. ‘directory’ for filesystem path completion)
SlashCommandInvocationResult: Result of invoking the slash command (text output, prompt to send to the agent, completion, or subcommand selection).
SlashCommandKind: Coarse command category for grouping and behavior: runtime built-in, skill-backed command, or SDK/client-owned command
SlashCommandSelectSubcommandResultKind: Select subcommand result discriminator
SlashCommandTextResultKind: Text result discriminator
SubagentSettingsEntryContextTier: Context tier override for matching subagents
TaskAgentInfoType: Task kind
TaskAgentProgressType: Progress kind
TaskExecutionMode: Whether task execution is synchronously awaited or managed in the background
TaskShellInfoAttachmentMode: Whether the shell runs inside a managed PTY session or as an independent background process
TaskShellInfoType: Task kind
TaskShellProgressType: Progress kind
TaskStatus: Current lifecycle status of the task
TokenAuthInfoType: SDK-side token authentication; the host configured the token directly via the SDK.
UIAutoModeSwitchResponse: User’s choice for auto-mode switching: yes (allow this turn), yes_always (allow + persist as setting), or no (decline).
UIElicitationArrayAnyOfFieldType: Type discriminator. Always “array”.
UIElicitationArrayEnumFieldItemsType: Type discriminator. Always “string”.
UIElicitationArrayEnumFieldType: Type discriminator. Always “array”.
UIElicitationResponseAction: The user’s response: accept (submitted), decline (rejected), or cancel (dismissed)
UIElicitationSchemaPropertyBooleanType: Type discriminator. Always “boolean”.
UIElicitationSchemaPropertyNumberType: Numeric type accepted by the field.
UIElicitationSchemaPropertyStringFormat: Optional format hint that constrains the accepted input.
UIElicitationSchemaPropertyStringType: Type discriminator. Always “string”.
UIElicitationSchemaType: Schema type indicator (always ‘object’)
UIElicitationStringEnumFieldType: Type discriminator. Always “string”.
UIElicitationStringOneOfFieldType: Type discriminator. Always “string”.
UIExitPlanModeAction: The action the user selected. Defaults to ‘autopilot’ when autoApproveEdits is true, otherwise ‘interactive’.
UISessionLimitsExhaustedResponseAction: User action selected for an exhausted session limit.
UserAuthInfoType: OAuth user authentication. The token itself is held in the runtime’s secret token store (keyed by host+login) and is NOT carried in this struct.
WorkspaceDiffFileChangeType: Type of change represented by this file diff.
WorkspaceDiffMode: Diff mode requested by the client.
WorkspaceSummaryHostType: Repository host type, if known
WorkspacesWorkspaceDetailsHostType: Allowed values for the WorkspacesWorkspaceDetailsHostType enumeration.

Type Aliases§

AccountGetAllUsersResult: List of all authenticated users
LlmInferenceHeaders: HTTP headers as a map from lowercased header name to a list of values. Multi-valued headers (e.g. Set-Cookie) preserve all values.
McpExecuteSamplingResult: MCP CreateMessageResult payload (with optional ‘tools’ extension), present when action=‘success’. Treated as opaque at the schema layer; consumers should construct/consume it per the MCP CreateMessageResult shape.
SessionMcpAppsCallToolResult: Standard MCP CallToolResult
UIElicitationResponseContent: The form values submitted by the user (present when action is ‘accept’)

Module rpc

Module rpc Copy item path

Modules§

Structs§

Enums§

Type Aliases§

Module rpc