wehub/simstudioai--sim
1
0
Fork 0
Code Issues Pull Requests Actions 1 Deployments Models Agent Notes Packages Projects Releases Wiki Activity
Files
d25d482dc25950f0ef5485d2ca60b36d5c6ca5b1
simstudioai--sim/apps/sim/lib/copilot/generated/tool-schemas-v1.ts
T
wehub-resource-sync d25d482dc2
CI / Migrate Dev DB (push) Has been skipped
Details
CI / Detect Version (push) Has been cancelled
Details
CI / Migrate DB (push) Has been cancelled
Details
CI / Build Dev ECR (./docker/app.Dockerfile, ECR_APP) (push) Has been cancelled
Details
CI / Build Dev ECR (./docker/db.Dockerfile, ECR_MIGRATIONS) (push) Has been cancelled
Details
CI / Build Dev ECR (./docker/pii.Dockerfile, ECR_PII) (push) Has been cancelled
Details
CI / Build Dev ECR (./docker/realtime.Dockerfile, ECR_REALTIME) (push) Has been cancelled
Details
CI / Deploy Trigger.dev (Dev) (push) Has been cancelled
Details
CI / Build AMD64 (./docker/app.Dockerfile, ECR_APP, ghcr.io/simstudioai/simstudio) (push) Has been cancelled
Details
CI / Build AMD64 (./docker/db.Dockerfile, ECR_MIGRATIONS, ghcr.io/simstudioai/migrations) (push) Has been cancelled
Details
CI / Build AMD64 (./docker/pii.Dockerfile, ECR_PII, ghcr.io/simstudioai/pii) (push) Has been cancelled
Details
CI / Build AMD64 (./docker/realtime.Dockerfile, ECR_REALTIME, ghcr.io/simstudioai/realtime) (push) Has been cancelled
Details
CI / Build ARM64 (GHCR Only) (./docker/app.Dockerfile, ghcr.io/simstudioai/simstudio) (push) Has been cancelled
Details
CI / Build ARM64 (GHCR Only) (./docker/db.Dockerfile, ghcr.io/simstudioai/migrations) (push) Has been cancelled
Details
CI / Build ARM64 (GHCR Only) (./docker/pii.Dockerfile, ghcr.io/simstudioai/pii) (push) Has been cancelled
Details
CI / Build ARM64 (GHCR Only) (./docker/realtime.Dockerfile, ghcr.io/simstudioai/realtime) (push) Has been cancelled
Details
CI / Create GHCR Manifests (ghcr.io/simstudioai/migrations) (push) Has been cancelled
Details
CI / Create GHCR Manifests (ghcr.io/simstudioai/pii) (push) Has been cancelled
Details
CI / Create GHCR Manifests (ghcr.io/simstudioai/realtime) (push) Has been cancelled
Details
CI / Create GHCR Manifests (ghcr.io/simstudioai/simstudio) (push) Has been cancelled
Details
CI / Check Docs Changes (push) Has been cancelled
Details
CI / Process Docs (push) Has been cancelled
Details
CI / Create GitHub Release (push) Has been cancelled
Details
CI / Test and Build (push) Has been cancelled
Details
Publish CLI Package / publish-npm (push) Has been cancelled
Details
Publish Python SDK / publish-pypi (push) Has been cancelled
Details
Publish TypeScript SDK / publish-npm (push) Has been cancelled
Details
chore: import upstream snapshot with attribution
2026-07-13 13:20:55 +08:00

4185 lines
143 KiB
TypeScript
Raw Blame History

// AUTO-GENERATED FILE. DO NOT EDIT.
// Generated from copilot/contracts/tool-catalog-v1.json
//
export type JsonSchema = unknown
export interface ToolRuntimeSchemaEntry {
parameters?: JsonSchema
resultSchema?: JsonSchema
}
export const TOOL_RUNTIME_SCHEMAS: Record<string, ToolRuntimeSchemaEntry> = {
agent: {
parameters: {
properties: {
request: {
description: 'What tool/skill/MCP action is needed.',
type: 'string',
},
},
required: ['request'],
type: 'object',
},
resultSchema: undefined,
},
auth: {
parameters: {
properties: {
request: {
description: 'What authentication/credential action is needed.',
type: 'string',
},
},
required: ['request'],
type: 'object',
},
resultSchema: undefined,
},
check_deployment_status: {
parameters: {
type: 'object',
properties: {
workflowId: {
type: 'string',
description: 'Workflow ID to check (defaults to current workflow)',
},
},
},
resultSchema: undefined,
},
complete_scheduled_task: {
parameters: {
type: 'object',
properties: {
jobId: {
type: 'string',
description: 'The ID of the scheduled task to mark as completed.',
},
},
required: ['jobId'],
},
resultSchema: undefined,
},
crawl_website: {
parameters: {
type: 'object',
properties: {
exclude_paths: {
type: 'array',
description: 'Skip URLs matching these patterns',
items: {
type: 'string',
},
},
include_paths: {
type: 'array',
description: 'Only crawl URLs matching these patterns',
items: {
type: 'string',
},
},
limit: {
type: 'number',
description: 'Maximum pages to crawl (default 10, max 50)',
},
max_depth: {
type: 'number',
description: 'How deep to follow links (default 2)',
},
url: {
type: 'string',
description: 'Starting URL to crawl from',
},
},
required: ['url'],
},
resultSchema: undefined,
},
create_file: {
parameters: {
type: 'object',
properties: {
contentType: {
type: 'string',
description:
'Optional MIME type override. Usually omit and let the system infer from the file extension.',
},
fileName: {
type: 'string',
description:
'Backward-compatible workspace filename. Prefer outputs.files[0].path for new calls.',
},
outputs: {
type: 'object',
description: 'Workspace file output declarations using canonical VFS paths.',
properties: {
files: {
type: 'array',
description:
'Files to create or overwrite. Parent folders must already exist for create mode.',
items: {
type: 'object',
properties: {
mimeType: {
type: 'string',
description: 'Optional MIME type override when inference is not enough.',
},
mode: {
type: 'string',
description: 'Create a new file or overwrite an existing file at path.',
enum: ['create', 'overwrite'],
},
path: {
type: 'string',
description: 'Canonical destination VFS path, e.g. "files/Reports/result.csv".',
},
},
required: ['path', 'mode'],
},
},
},
},
},
},
resultSchema: {
type: 'object',
properties: {
data: {
type: 'object',
description:
'Contains id (internal file ID), name, and vfsPath. Use vfsPath for follow-up file tools.',
},
message: {
type: 'string',
description: 'Human-readable outcome.',
},
success: {
type: 'boolean',
description: 'Whether the file was created.',
},
},
required: ['success', 'message'],
},
},
create_file_folder: {
parameters: {
type: 'object',
properties: {
path: {
type: 'string',
description:
'Canonical folder VFS path to create, e.g. "files/Images" or "files/Reports/2026".',
},
workspaceId: {
type: 'string',
description: 'Optional workspace ID. Defaults to the current workspace.',
},
},
required: ['path'],
},
resultSchema: undefined,
},
create_workflow: {
parameters: {
type: 'object',
properties: {
description: {
type: 'string',
description: 'Optional workflow description.',
},
folderId: {
type: 'string',
description: 'Optional folder ID.',
},
name: {
type: 'string',
description: 'Workflow name.',
},
workspaceId: {
type: 'string',
description: 'Optional workspace ID.',
},
},
required: ['name'],
},
resultSchema: undefined,
},
create_workspace_mcp_server: {
parameters: {
type: 'object',
properties: {
description: {
type: 'string',
description: 'Optional description for the server',
},
isPublic: {
type: 'boolean',
description: 'Whether the workflow MCP server is publicly accessible',
},
name: {
type: 'string',
description: 'Required: server name',
},
workflowIds: {
type: 'array',
description: 'Optional deployed workflow IDs to publish as tools on the new server',
items: {
type: 'string',
},
},
workspaceId: {
type: 'string',
description:
'Workspace ID. Required when no current workspace context is available, such as headless MCP calls.',
},
},
required: ['name'],
},
resultSchema: undefined,
},
delete_file: {
parameters: {
type: 'object',
properties: {
paths: {
type: 'array',
description:
'Canonical workspace file VFS paths to delete, e.g. ["files/Reports/draft.md"].',
items: {
type: 'string',
},
},
},
required: ['paths'],
},
resultSchema: {
type: 'object',
properties: {
message: {
type: 'string',
description: 'Human-readable outcome.',
},
success: {
type: 'boolean',
description: 'Whether the delete succeeded.',
},
},
required: ['success', 'message'],
},
},
delete_file_folder: {
parameters: {
type: 'object',
properties: {
paths: {
type: 'array',
description: 'Canonical folder VFS paths to delete, e.g. ["files/Archive"].',
items: {
type: 'string',
},
},
},
required: ['paths'],
},
resultSchema: undefined,
},
delete_workflow: {
parameters: {
type: 'object',
properties: {
workflowIds: {
type: 'array',
description: 'The workflow IDs to delete.',
items: {
type: 'string',
},
},
},
required: ['workflowIds'],
},
resultSchema: undefined,
},
delete_workspace_mcp_server: {
parameters: {
type: 'object',
properties: {
serverId: {
type: 'string',
description: 'Required: the MCP server ID to delete',
},
},
required: ['serverId'],
},
resultSchema: undefined,
},
deploy: {
parameters: {
properties: {
request: {
description:
'Detailed deployment instructions. Include the deployment type and ALL user-specified options: identifier, title, description, authType, password, allowedEmails, welcomeMessage, outputConfigs (block outputs to display).',
type: 'string',
},
},
required: ['request'],
type: 'object',
},
resultSchema: undefined,
},
deploy_api: {
parameters: {
type: 'object',
properties: {
action: {
type: 'string',
description: 'Whether to deploy or undeploy the API endpoint',
enum: ['deploy', 'undeploy'],
default: 'deploy',
},
versionDescription: {
type: 'string',
description:
'REQUIRED when action is "deploy": a concise (1-3 sentence) description of what changed in this deployment version, e.g. "Adds Slack failure alert and retries on the HTTP block". If unsure what changed, call diff_workflows(ref1: "live", ref2: "draft") first. Ignored for undeploy.',
},
versionName: {
type: 'string',
description:
'REQUIRED when action is "deploy": a short human-readable name/label for this deployment version (shown in the deployment history), e.g. "v2 pricing" or "Add Slack alerts". Ignored for undeploy.',
},
workflowId: {
type: 'string',
description: 'Workflow ID to deploy (required in workspace context)',
},
},
},
resultSchema: {
type: 'object',
properties: {
apiEndpoint: {
type: 'string',
description: 'Canonical workflow execution endpoint.',
},
baseUrl: {
type: 'string',
description: 'Base URL used to construct deployment URLs.',
},
deployedAt: {
type: 'string',
description: 'Deployment timestamp when the workflow is deployed.',
},
deploymentConfig: {
type: 'object',
description:
'Structured deployment configuration keyed by surface name. For API deploys this includes endpoint, auth, and sync/stream/async mode details.',
},
deploymentStatus: {
type: 'object',
description:
'Structured per-surface deployment status keyed by surface name, such as api.',
},
deploymentType: {
type: 'string',
description:
'Deployment surface this result describes. For deploy_api and redeploy this is always "api".',
},
examples: {
type: 'object',
description:
'Invocation examples keyed by surface name. For API deploys this includes curl examples for sync, stream, async, and polling.',
},
isDeployed: {
type: 'boolean',
description: 'Whether the workflow API is currently deployed after this tool call.',
},
version: {
type: 'number',
description: 'Deployment version for the current API deployment.',
},
workflowId: {
type: 'string',
description: 'Workflow ID that was deployed or undeployed.',
},
},
required: [
'workflowId',
'isDeployed',
'deploymentType',
'deploymentStatus',
'deploymentConfig',
'examples',
],
},
},
deploy_chat: {
parameters: {
type: 'object',
properties: {
action: {
type: 'string',
description: 'Whether to deploy or undeploy the chat interface',
enum: ['deploy', 'undeploy'],
default: 'deploy',
},
allowedEmails: {
type: 'array',
description: 'List of allowed emails/domains for email or SSO auth',
items: {
type: 'string',
},
},
authType: {
type: 'string',
description: 'Authentication type: public, password, email, or sso',
enum: ['public', 'password', 'email', 'sso'],
default: 'public',
},
description: {
type: 'string',
description: 'Optional chat-facing description shown on the chat page',
},
identifier: {
type: 'string',
description: 'URL slug for the chat (lowercase letters, numbers, hyphens only)',
},
outputConfigs: {
type: 'array',
description: 'Output configurations specifying which block outputs to display in chat',
items: {
type: 'object',
properties: {
blockId: {
type: 'string',
description: 'The block UUID',
},
path: {
type: 'string',
description:
'The output path (e.g. `content` for an agent; structured fields are top-level paths). Call get_block_outputs for real paths.',
},
},
required: ['blockId', 'path'],
},
},
password: {
type: 'string',
description: 'Password for password-protected chats',
},
title: {
type: 'string',
description: 'Display title for the chat interface',
},
versionDescription: {
type: 'string',
description:
'REQUIRED when action is "deploy": a concise (1-3 sentence) description of what changed in this deployment version (distinct from the chat-facing description). If unsure what changed, call diff_workflows(ref1: "live", ref2: "draft") first. Ignored for undeploy.',
},
versionName: {
type: 'string',
description:
'REQUIRED when action is "deploy": a short human-readable name/label for this deployment version (distinct from the chat title; shown in deployment history). Ignored for undeploy.',
},
welcomeMessage: {
type: 'string',
description: 'Welcome message shown to users',
},
workflowId: {
type: 'string',
description: 'Workflow ID to deploy (required in workspace context)',
},
},
},
resultSchema: {
type: 'object',
properties: {
action: {
type: 'string',
description: 'Action performed by the tool, such as "deploy" or "undeploy".',
},
apiEndpoint: {
type: 'string',
description: 'Paired workflow execution endpoint used by the chat deployment.',
},
baseUrl: {
type: 'string',
description: 'Base URL used to construct deployment URLs.',
},
chatUrl: {
type: 'string',
description: 'Shareable chat URL when the chat surface is deployed.',
},
deployedAt: {
type: 'string',
description: 'Deployment timestamp for the underlying workflow deployment.',
},
deploymentConfig: {
type: 'object',
description:
'Structured deployment configuration keyed by surface name. Includes chat settings and the paired API invocation configuration.',
},
deploymentStatus: {
type: 'object',
description:
'Structured per-surface deployment status keyed by surface name, including api and chat.',
},
deploymentType: {
type: 'string',
description:
'Deployment surface this result describes. For deploy_chat this is always "chat".',
},
examples: {
type: 'object',
description:
'Invocation examples keyed by surface name. Includes chat access details and API curl examples.',
},
identifier: {
type: 'string',
description: 'Chat identifier or slug.',
},
isChatDeployed: {
type: 'boolean',
description: 'Whether the chat surface is deployed after this tool call.',
},
isDeployed: {
type: 'boolean',
description: 'Whether the paired API surface remains deployed after this tool call.',
},
success: {
type: 'boolean',
description: 'Whether the deploy_chat action completed successfully.',
},
version: {
type: 'number',
description: 'Deployment version for the underlying workflow deployment.',
},
workflowId: {
type: 'string',
description: 'Workflow ID associated with the chat deployment.',
},
},
required: [
'workflowId',
'success',
'action',
'isDeployed',
'isChatDeployed',
'deploymentType',
'deploymentStatus',
'deploymentConfig',
'examples',
],
},
},
deploy_custom_block: {
parameters: {
type: 'object',
properties: {
action: {
type: 'string',
description: 'Whether to publish (deploy) or unpublish (undeploy) the custom block',
enum: ['deploy', 'undeploy'],
default: 'deploy',
},
description: {
type: 'string',
description: 'Short description shown in the block picker, max 280 characters',
},
exposedOutputs: {
type: 'array',
description:
"Outputs the block exposes, each mapping a child block output path to a friendly name (use get_block_outputs for valid paths). Omit to expose the terminal block's whole result",
items: {
type: 'object',
properties: {
blockId: {
type: 'string',
description: 'Block UUID inside the workflow',
},
name: {
type: 'string',
description: 'Friendly output name shown on the block',
},
path: {
type: 'string',
description:
"Dot-path into that block's output (from get_block_outputs relativeOutputs)",
},
},
required: ['blockId', 'path', 'name'],
},
},
iconUrl: {
type: 'string',
description:
'Optional icon image for the block: a workspace file VFS path (e.g. "files/icon.png", copied into public icon storage at publish) or an external image URL. Omit to use the organization\'s default icon',
},
inputs: {
type: 'array',
description:
"Optional per-input placeholder overrides. Input names and types are derived from the workflow's input trigger and cannot be changed here",
items: {
type: 'object',
properties: {
id: {
type: 'string',
description: 'Stable id of the input trigger field',
},
placeholder: {
type: 'string',
description: "Placeholder text shown in the block's input field",
},
},
required: ['id'],
},
},
name: {
type: 'string',
description:
'Display name for the block, max 60 characters. When republishing an existing block, pass the current name to keep it or a new name to rename.',
},
workflowId: {
type: 'string',
description: 'Workflow ID (defaults to active workflow)',
},
},
required: ['name'],
},
resultSchema: {
type: 'object',
properties: {
action: {
type: 'string',
description: 'Action performed by the tool, such as "deploy" or "undeploy".',
},
blockId: {
type: 'string',
description: 'Custom block record ID.',
},
blockType: {
type: 'string',
description: 'Stable block type slug (custom_block_*) used in workflow state.',
},
deploymentConfig: {
type: 'object',
description:
"Structured deployment configuration keyed by surface name. Includes the block's type, name, description, icon, derived input fields, and exposed outputs.",
},
deploymentStatus: {
type: 'object',
description:
'Structured per-surface deployment status keyed by surface name, including customBlock and the underlying api surface when applicable.',
},
deploymentType: {
type: 'string',
description:
'Deployment surface this result describes. For deploy_custom_block this is always "custom_block".',
},
isDeployed: {
type: 'boolean',
description: 'Whether the custom block is published after this tool call.',
},
name: {
type: 'string',
description: 'Display name of the custom block.',
},
removed: {
type: 'boolean',
description: 'Whether the custom block was unpublished during an undeploy action.',
},
updated: {
type: 'boolean',
description: 'Whether an existing custom block was updated instead of created.',
},
workflowId: {
type: 'string',
description: 'Workflow ID the custom block is bound to.',
},
},
required: ['deploymentType', 'deploymentStatus'],
},
},
deploy_mcp: {
parameters: {
type: 'object',
properties: {
parameterDescriptions: {
type: 'array',
description: 'Array of parameter descriptions for the tool',
items: {
type: 'object',
properties: {
description: {
type: 'string',
description: 'Parameter description',
},
name: {
type: 'string',
description: 'Parameter name',
},
},
required: ['name', 'description'],
},
},
serverId: {
type: 'string',
description: 'Required: server ID from list_workspace_mcp_servers',
},
toolDescription: {
type: 'string',
description: 'Description for the MCP tool',
},
toolName: {
type: 'string',
description: 'Name for the MCP tool (defaults to workflow name)',
},
workflowId: {
type: 'string',
description: 'Workflow ID (defaults to active workflow)',
},
},
required: ['serverId'],
},
resultSchema: {
type: 'object',
properties: {
action: {
type: 'string',
description: 'Action performed by the tool, such as "deploy" or "undeploy".',
},
apiEndpoint: {
type: 'string',
description: 'Underlying workflow API endpoint associated with the MCP tool.',
},
baseUrl: {
type: 'string',
description: 'Base URL used to construct deployment URLs.',
},
deploymentConfig: {
type: 'object',
description:
'Structured deployment configuration keyed by surface name. Includes MCP server, tool, auth, and parameter schema details.',
},
deploymentStatus: {
type: 'object',
description:
'Structured per-surface deployment status keyed by surface name, including mcp and the underlying api surface when applicable.',
},
deploymentType: {
type: 'string',
description:
'Deployment surface this result describes. For deploy_mcp this is always "mcp".',
},
examples: {
type: 'object',
description:
'Setup examples keyed by surface name. Includes ready-to-paste config snippets for supported MCP clients.',
},
mcpServerUrl: {
type: 'string',
description: 'HTTP MCP server URL to configure in clients.',
},
removed: {
type: 'boolean',
description: 'Whether the MCP deployment was removed during an undeploy action.',
},
serverId: {
type: 'string',
description: 'Workspace MCP server ID.',
},
serverName: {
type: 'string',
description: 'Workspace MCP server name.',
},
toolDescription: {
type: 'string',
description: 'MCP tool description exposed on the server.',
},
toolId: {
type: 'string',
description: 'MCP tool ID when deployed.',
},
toolName: {
type: 'string',
description: 'MCP tool name exposed on the server.',
},
updated: {
type: 'boolean',
description: 'Whether an existing MCP tool deployment was updated instead of created.',
},
workflowId: {
type: 'string',
description: 'Workflow ID associated with the MCP deployment.',
},
},
required: ['deploymentType', 'deploymentStatus'],
},
},
diff_workflows: {
parameters: {
type: 'object',
properties: {
ref1: {
type: 'string',
description:
'Base side (string): a version number (e.g. "3"), "live" (active deployment), or "draft" (current editor state).',
},
ref2: {
type: 'string',
description:
'Target side (string): a version number (e.g. "4"), "live" (active deployment), or "draft" (current editor state).',
},
workflowId: {
type: 'string',
description:
'Optional workflow ID. If not provided, uses the current workflow in context.',
},
},
required: ['ref1', 'ref2'],
},
resultSchema: undefined,
},
download_to_workspace_file: {
parameters: {
type: 'object',
properties: {
fileName: {
type: 'string',
description:
'Backward-compatible workspace file name. Prefer outputs.files[0].path for new calls.',
},
outputs: {
type: 'object',
description: 'Workspace file output declarations using canonical VFS paths.',
properties: {
files: {
type: 'array',
description:
'Files to create or overwrite. Parent folders must already exist for create mode.',
items: {
type: 'object',
properties: {
mimeType: {
type: 'string',
description: 'Optional MIME type override when inference is not enough.',
},
mode: {
type: 'string',
description: 'Create a new file or overwrite an existing file at path.',
enum: ['create', 'overwrite'],
},
path: {
type: 'string',
description: 'Canonical destination VFS path, e.g. "files/Reports/result.csv".',
},
},
required: ['path', 'mode'],
},
},
},
},
url: {
type: 'string',
description:
'Direct URL of the file to download, such as an image CDN URL ending in .png or .jpg',
},
},
required: ['url'],
},
resultSchema: undefined,
},
edit_content: {
parameters: {
type: 'object',
properties: {
content: {
type: 'string',
description:
'The text content to write. For append: text to append. For update: full replacement text. For patch with search_replace: the replacement text. For patch with anchored: the insert/replacement text.',
},
},
required: ['content'],
},
resultSchema: {
type: 'object',
properties: {
data: {
type: 'object',
description:
'Optional operation metadata such as file id, file name, size, and content type.',
},
message: {
type: 'string',
description: 'Human-readable summary of the outcome.',
},
success: {
type: 'boolean',
description: 'Whether the content was applied successfully.',
},
},
required: ['success', 'message'],
},
},
edit_workflow: {
parameters: {
type: 'object',
properties: {
operations: {
type: 'array',
description: 'Array of edit operations',
items: {
type: 'object',
properties: {
block_id: {
type: 'string',
description:
'Block ID for the operation. For add operations, this will be the desired ID for the new block.',
},
operation_type: {
type: 'string',
description: 'Type of operation to perform',
enum: ['add', 'edit', 'delete', 'insert_into_subflow', 'extract_from_subflow'],
},
params: {
type: 'object',
description:
'Parameters for the operation (optional).\nFor edit: {"inputs": {"temperature": 0.5}} NOT {"subBlocks": {"temperature": {"value": 0.5}}}\nFor add: {"type": "agent", "name": "My Agent", "inputs": {"model": "<model-id from agent.json>"}}\nFor delete: omit params entirely (none needed)',
},
},
required: ['operation_type', 'block_id'],
},
},
workflowId: {
type: 'string',
description:
'Optional workflow ID to edit. If not provided, uses the current workflow in context.',
},
},
required: ['operations'],
},
resultSchema: undefined,
},
enrichment_run: {
parameters: {
type: 'object',
properties: {
enrichmentId: {
type: 'string',
description:
"Which enrichment to run. Discover the full set and each one's inputs/outputs via user_table.list_enrichments.",
enum: [
'work-email',
'phone-number',
'company-domain',
'company-info',
'email-verification',
],
},
inputs: {
type: 'object',
description:
'Map of the enrichment\'s input id → value, e.g. { "fullName": "Jane Doe", "companyDomain": "acme.com" }. Provide a value for every required input.',
},
},
required: ['enrichmentId', 'inputs'],
},
resultSchema: {
type: 'object',
properties: {
matched: {
type: 'boolean',
description: 'True when a provider returned a non-empty result.',
},
provider: {
type: 'string',
description:
'Internal label of the provider that produced the result (billing/diagnostics only — do NOT surface it to the user), or null on no match.',
},
result: {
type: 'object',
description: 'Mapped output values from the winning provider (empty object on no match).',
},
},
required: ['matched', 'result'],
},
},
ffmpeg: {
parameters: {
type: 'object',
properties: {
aspectRatio: {
type: 'string',
description: 'Target aspect ratio for scale_pad, e.g. 9:16, 16:9, 1:1.',
},
end: {
type: 'number',
description: 'End time in seconds (trim).',
},
format: {
type: 'string',
description: 'Target format/extension for convert (e.g. mp4, mp3, wav, gif).',
},
height: {
type: 'number',
description: 'Target height in pixels (scale_pad).',
},
inputs: {
type: 'object',
description:
'Workspace resources to mount into the sandbox. Copy paths verbatim from glob/read/grep output — they are percent-encoded per segment (spaces are %20, an in-name slash is %2F; parentheses and dots stay literal). Both the encoded path and the plain name resolve, so copy the returned path exactly rather than retyping or decoding it.',
properties: {
directories: {
type: 'array',
description:
'Workspace folders to mount recursively into the sandbox, including nested files and empty folders.',
items: {
type: 'object',
properties: {
path: {
type: 'string',
description:
'Canonical VFS folder path, e.g. "files/Reports". By default this mounts at "/home/user/{path}".',
},
sandboxPath: {
type: 'string',
description:
'Optional full sandbox directory path override. Omit to mount at /home/user/{path}.',
},
},
required: ['path'],
},
},
files: {
type: 'array',
description: 'Workspace files to mount into the sandbox.',
items: {
type: 'object',
properties: {
path: {
type: 'string',
description:
'Canonical VFS file path, e.g. "files/Reports/sales.csv". By default this mounts at "/home/user/{path}".',
},
sandboxPath: {
type: 'string',
description:
'Full sandbox path to mount at, e.g. /home/user/inputs/data.csv. STRONGLY RECOMMENDED whenever the file name has spaces or special characters: the default mount path is the percent-ENCODED canonical path (e.g. /home/user/files/Q4%20Sales%20(Final).csv), which code using the human-readable name will not find. Set a simple sandboxPath and read exactly that.',
},
},
required: ['path'],
},
},
tables: {
type: 'array',
description: 'Workspace tables to mount as CSV files.',
items: {
type: 'object',
properties: {
path: {
type: 'string',
description: 'Canonical VFS table path when available.',
},
sandboxPath: {
type: 'string',
description: 'Optional full sandbox path for the mounted CSV.',
},
tableId: {
type: 'string',
description: 'Workspace table ID.',
},
},
},
},
},
},
loopToVideo: {
type: 'boolean',
description: 'For overlay_audio, loop or trim the audio to match the video length.',
},
musicVolume: {
type: 'number',
description: 'Volume multiplier for the background music track in mix_audio (e.g. 0.3).',
},
operation: {
type: 'string',
description: 'The FFmpeg operation to run.',
enum: [
'overlay_audio',
'mix_audio',
'concat',
'trim',
'scale_pad',
'overlay_image',
'add_text',
'fade',
'extract_audio',
'convert',
'thumbnail',
'probe',
],
},
outputs: {
type: 'object',
description:
'Workspace files to create or overwrite from returned code results or sandbox-created files.',
properties: {
files: {
type: 'array',
description: 'File outputs. Parent folders must already exist for create mode.',
items: {
type: 'object',
properties: {
format: {
type: 'string',
description: 'Optional serialization format for returned values.',
enum: ['json', 'csv', 'txt', 'md', 'html'],
},
mimeType: {
type: 'string',
description: 'Optional MIME type override when inference is not enough.',
},
mode: {
type: 'string',
description: 'Create a new file or overwrite an existing file at path.',
enum: ['create', 'overwrite'],
},
path: {
type: 'string',
description: 'Canonical destination VFS path, e.g. "files/Reports/chart.png".',
},
sandboxPath: {
type: 'string',
description:
'Optional full path to a file created inside the sandbox. Omit to save the code return value.',
},
},
required: ['path', 'mode'],
},
},
},
},
position: {
type: 'string',
description: 'Placement for add_text / overlay_image.',
enum: ['top', 'center', 'bottom', 'top-left', 'top-right', 'bottom-left', 'bottom-right'],
},
start: {
type: 'number',
description: 'Start time in seconds (trim, thumbnail, fade).',
},
text: {
type: 'string',
description: 'Text to burn in for add_text.',
},
volume: {
type: 'number',
description: 'Volume multiplier for the primary track (mix_audio / overlay_audio).',
},
width: {
type: 'number',
description: 'Target width in pixels (scale_pad).',
},
},
required: ['operation', 'inputs'],
},
resultSchema: undefined,
},
file: {
parameters: {
properties: {
prompt: {
description:
"Optional brief instruction (one short sentence) to scope the task. The agent inherits the full conversation history — do NOT restate or rewrite conversation content, only add scoping the history doesn't convey.",
type: 'string',
},
},
type: 'object',
},
resultSchema: undefined,
},
function_execute: {
parameters: {
type: 'object',
properties: {
code: {
type: 'string',
description:
'Code to execute. For JS: raw statements auto-wrapped in async context. For Python: full script. For shell: bash script with access to pre-installed CLI tools and workspace env vars as $VAR_NAME.',
},
inputs: {
type: 'object',
description:
'Workspace resources to mount into the sandbox. Copy paths verbatim from glob/read/grep output — they are percent-encoded per segment (spaces are %20, an in-name slash is %2F; parentheses and dots stay literal). Both the encoded path and the plain name resolve, so copy the returned path exactly rather than retyping or decoding it.',
properties: {
directories: {
type: 'array',
description:
'Workspace folders to mount recursively into the sandbox, including nested files and empty folders.',
items: {
type: 'object',
properties: {
path: {
type: 'string',
description:
'Canonical VFS folder path, e.g. "files/Reports". By default this mounts at "/home/user/{path}".',
},
sandboxPath: {
type: 'string',
description:
'Optional full sandbox directory path override. Omit to mount at /home/user/{path}.',
},
},
required: ['path'],
},
},
files: {
type: 'array',
description: 'Workspace files to mount into the sandbox.',
items: {
type: 'object',
properties: {
path: {
type: 'string',
description:
'Canonical VFS file path, e.g. "files/Reports/sales.csv". By default this mounts at "/home/user/{path}".',
},
sandboxPath: {
type: 'string',
description:
'Full sandbox path to mount at, e.g. /home/user/inputs/data.csv. STRONGLY RECOMMENDED whenever the file name has spaces or special characters: the default mount path is the percent-ENCODED canonical path (e.g. /home/user/files/Q4%20Sales%20(Final).csv), which code using the human-readable name will not find. Set a simple sandboxPath and read exactly that.',
},
},
required: ['path'],
},
},
tables: {
type: 'array',
description: 'Workspace tables to mount as CSV files.',
items: {
type: 'object',
properties: {
path: {
type: 'string',
description: 'Canonical VFS table path when available.',
},
sandboxPath: {
type: 'string',
description: 'Optional full sandbox path for the mounted CSV.',
},
tableId: {
type: 'string',
description: 'Workspace table ID.',
},
},
},
},
},
},
language: {
type: 'string',
description: 'Execution language.',
enum: ['javascript', 'python', 'shell'],
},
outputTable: {
type: 'string',
description:
'Table ID to overwrite with the code\'s return value. Code MUST return an array of objects where keys match column names. All existing rows are replaced. Example: "tbl_abc123"',
},
outputs: {
type: 'object',
description:
'Workspace files to create or overwrite from returned code results or sandbox-created files.',
properties: {
files: {
type: 'array',
description: 'File outputs. Parent folders must already exist for create mode.',
items: {
type: 'object',
properties: {
format: {
type: 'string',
description: 'Optional serialization format for returned values.',
enum: ['json', 'csv', 'txt', 'md', 'html'],
},
mimeType: {
type: 'string',
description: 'Optional MIME type override when inference is not enough.',
},
mode: {
type: 'string',
description: 'Create a new file or overwrite an existing file at path.',
enum: ['create', 'overwrite'],
},
path: {
type: 'string',
description: 'Canonical destination VFS path, e.g. "files/Reports/chart.png".',
},
sandboxPath: {
type: 'string',
description:
'Optional full path to a file created inside the sandbox. Omit to save the code return value.',
},
},
required: ['path', 'mode'],
},
},
},
},
title: {
type: 'string',
description:
'Short user-visible label for this execution, e.g. "Clean customer CSV", "Revenue chart", or "Query GitHub issues".',
},
},
required: ['code'],
},
resultSchema: undefined,
},
generate_api_key: {
parameters: {
type: 'object',
properties: {
name: {
type: 'string',
description:
"A descriptive name for the API key (e.g., 'production-key', 'dev-testing').",
},
workspaceId: {
type: 'string',
description: "Optional workspace ID. Defaults to user's default workspace.",
},
},
required: ['name'],
},
resultSchema: undefined,
},
generate_audio: {
parameters: {
type: 'object',
properties: {
duration: {
type: 'number',
description:
'Approximate duration in seconds for sfx (and music models that support it). MiniMax music ignores this — fit music to a video with the ffmpeg tool instead.',
},
inputs: {
type: 'object',
description:
'Workspace resources to mount into the sandbox. Copy paths verbatim from glob/read/grep output — they are percent-encoded per segment (spaces are %20, an in-name slash is %2F; parentheses and dots stay literal). Both the encoded path and the plain name resolve, so copy the returned path exactly rather than retyping or decoding it.',
properties: {
directories: {
type: 'array',
description:
'Workspace folders to mount recursively into the sandbox, including nested files and empty folders.',
items: {
type: 'object',
properties: {
path: {
type: 'string',
description:
'Canonical VFS folder path, e.g. "files/Reports". By default this mounts at "/home/user/{path}".',
},
sandboxPath: {
type: 'string',
description:
'Optional full sandbox directory path override. Omit to mount at /home/user/{path}.',
},
},
required: ['path'],
},
},
files: {
type: 'array',
description: 'Workspace files to mount into the sandbox.',
items: {
type: 'object',
properties: {
path: {
type: 'string',
description:
'Canonical VFS file path, e.g. "files/Reports/sales.csv". By default this mounts at "/home/user/{path}".',
},
sandboxPath: {
type: 'string',
description:
'Full sandbox path to mount at, e.g. /home/user/inputs/data.csv. STRONGLY RECOMMENDED whenever the file name has spaces or special characters: the default mount path is the percent-ENCODED canonical path (e.g. /home/user/files/Q4%20Sales%20(Final).csv), which code using the human-readable name will not find. Set a simple sandboxPath and read exactly that.',
},
},
required: ['path'],
},
},
tables: {
type: 'array',
description: 'Workspace tables to mount as CSV files.',
items: {
type: 'object',
properties: {
path: {
type: 'string',
description: 'Canonical VFS table path when available.',
},
sandboxPath: {
type: 'string',
description: 'Optional full sandbox path for the mounted CSV.',
},
tableId: {
type: 'string',
description: 'Workspace table ID.',
},
},
},
},
},
},
instrumental: {
type: 'boolean',
description:
'For music: true = instrumental, no vocals (default); false = a song with vocals.',
},
lyrics: {
type: 'string',
description:
'For music with vocals: the lyrics to sing (optional; supports [Verse]/[Chorus] tags). Setting this implies instrumental=false.',
},
model: {
type: 'string',
description:
'Optional model override for the selected type (e.g. fal-ai/elevenlabs/tts/eleven-v3 for speech).',
},
outputs: {
type: 'object',
description:
'Workspace files to create or overwrite from returned code results or sandbox-created files.',
properties: {
files: {
type: 'array',
description: 'File outputs. Parent folders must already exist for create mode.',
items: {
type: 'object',
properties: {
format: {
type: 'string',
description: 'Optional serialization format for returned values.',
enum: ['json', 'csv', 'txt', 'md', 'html'],
},
mimeType: {
type: 'string',
description: 'Optional MIME type override when inference is not enough.',
},
mode: {
type: 'string',
description: 'Create a new file or overwrite an existing file at path.',
enum: ['create', 'overwrite'],
},
path: {
type: 'string',
description: 'Canonical destination VFS path, e.g. "files/Reports/chart.png".',
},
sandboxPath: {
type: 'string',
description:
'Optional full path to a file created inside the sandbox. Omit to save the code return value.',
},
},
required: ['path', 'mode'],
},
},
},
},
prompt: {
type: 'string',
description:
'For speech: the text to speak (may include expressive tags). For music/sfx: a description of the audio to generate.',
},
type: {
type: 'string',
description: 'Kind of audio to generate. Defaults to speech.',
enum: ['speech', 'music', 'sfx'],
},
voice: {
type: 'string',
description: 'Optional voice name or id for speech.',
},
},
required: ['prompt'],
},
resultSchema: undefined,
},
generate_image: {
parameters: {
type: 'object',
properties: {
aspectRatio: {
type: 'string',
description: 'Aspect ratio for the generated image.',
enum: ['1:1', '16:9', '9:16', '4:3', '3:4'],
},
inputs: {
type: 'object',
description:
'Workspace resources to mount into the sandbox. Copy paths verbatim from glob/read/grep output — they are percent-encoded per segment (spaces are %20, an in-name slash is %2F; parentheses and dots stay literal). Both the encoded path and the plain name resolve, so copy the returned path exactly rather than retyping or decoding it.',
properties: {
directories: {
type: 'array',
description:
'Workspace folders to mount recursively into the sandbox, including nested files and empty folders.',
items: {
type: 'object',
properties: {
path: {
type: 'string',
description:
'Canonical VFS folder path, e.g. "files/Reports". By default this mounts at "/home/user/{path}".',
},
sandboxPath: {
type: 'string',
description:
'Optional full sandbox directory path override. Omit to mount at /home/user/{path}.',
},
},
required: ['path'],
},
},
files: {
type: 'array',
description: 'Workspace files to mount into the sandbox.',
items: {
type: 'object',
properties: {
path: {
type: 'string',
description:
'Canonical VFS file path, e.g. "files/Reports/sales.csv". By default this mounts at "/home/user/{path}".',
},
sandboxPath: {
type: 'string',
description:
'Full sandbox path to mount at, e.g. /home/user/inputs/data.csv. STRONGLY RECOMMENDED whenever the file name has spaces or special characters: the default mount path is the percent-ENCODED canonical path (e.g. /home/user/files/Q4%20Sales%20(Final).csv), which code using the human-readable name will not find. Set a simple sandboxPath and read exactly that.',
},
},
required: ['path'],
},
},
tables: {
type: 'array',
description: 'Workspace tables to mount as CSV files.',
items: {
type: 'object',
properties: {
path: {
type: 'string',
description: 'Canonical VFS table path when available.',
},
sandboxPath: {
type: 'string',
description: 'Optional full sandbox path for the mounted CSV.',
},
tableId: {
type: 'string',
description: 'Workspace table ID.',
},
},
},
},
},
},
outputs: {
type: 'object',
description:
'Workspace files to create or overwrite from returned code results or sandbox-created files.',
properties: {
files: {
type: 'array',
description: 'File outputs. Parent folders must already exist for create mode.',
items: {
type: 'object',
properties: {
format: {
type: 'string',
description: 'Optional serialization format for returned values.',
enum: ['json', 'csv', 'txt', 'md', 'html'],
},
mimeType: {
type: 'string',
description: 'Optional MIME type override when inference is not enough.',
},
mode: {
type: 'string',
description: 'Create a new file or overwrite an existing file at path.',
enum: ['create', 'overwrite'],
},
path: {
type: 'string',
description: 'Canonical destination VFS path, e.g. "files/Reports/chart.png".',
},
sandboxPath: {
type: 'string',
description:
'Optional full path to a file created inside the sandbox. Omit to save the code return value.',
},
},
required: ['path', 'mode'],
},
},
},
},
prompt: {
type: 'string',
description:
'Detailed text description of the image to generate, or editing instructions when editing the image(s) passed in `inputs.files`.',
},
},
required: ['prompt'],
},
resultSchema: undefined,
},
generate_video: {
parameters: {
type: 'object',
properties: {
aspectRatio: {
type: 'string',
description: 'Aspect ratio for the video (model-dependent).',
enum: ['16:9', '9:16', '1:1'],
},
duration: {
type: 'number',
description: 'Clip duration in seconds (model-dependent; e.g. 4, 6, 8).',
},
generateAudio: {
type: 'boolean',
description:
"Toggle Veo's native audio (dialogue/SFX/ambience/music generated from the prompt). Default true. Set false when you will add your own voiceover/music via the ffmpeg tool.",
},
inputs: {
type: 'object',
description:
'Workspace resources to mount into the sandbox. Copy paths verbatim from glob/read/grep output — they are percent-encoded per segment (spaces are %20, an in-name slash is %2F; parentheses and dots stay literal). Both the encoded path and the plain name resolve, so copy the returned path exactly rather than retyping or decoding it.',
properties: {
directories: {
type: 'array',
description:
'Workspace folders to mount recursively into the sandbox, including nested files and empty folders.',
items: {
type: 'object',
properties: {
path: {
type: 'string',
description:
'Canonical VFS folder path, e.g. "files/Reports". By default this mounts at "/home/user/{path}".',
},
sandboxPath: {
type: 'string',
description:
'Optional full sandbox directory path override. Omit to mount at /home/user/{path}.',
},
},
required: ['path'],
},
},
files: {
type: 'array',
description: 'Workspace files to mount into the sandbox.',
items: {
type: 'object',
properties: {
path: {
type: 'string',
description:
'Canonical VFS file path, e.g. "files/Reports/sales.csv". By default this mounts at "/home/user/{path}".',
},
sandboxPath: {
type: 'string',
description:
'Full sandbox path to mount at, e.g. /home/user/inputs/data.csv. STRONGLY RECOMMENDED whenever the file name has spaces or special characters: the default mount path is the percent-ENCODED canonical path (e.g. /home/user/files/Q4%20Sales%20(Final).csv), which code using the human-readable name will not find. Set a simple sandboxPath and read exactly that.',
},
},
required: ['path'],
},
},
tables: {
type: 'array',
description: 'Workspace tables to mount as CSV files.',
items: {
type: 'object',
properties: {
path: {
type: 'string',
description: 'Canonical VFS table path when available.',
},
sandboxPath: {
type: 'string',
description: 'Optional full sandbox path for the mounted CSV.',
},
tableId: {
type: 'string',
description: 'Workspace table ID.',
},
},
},
},
},
},
model: {
type: 'string',
description:
"Optional model override, keyed to the video's goal: veo-3.1-lite (prototype/quick test, cheapest), veo-3.1-fast (reasonable draft — default, good video), veo-3.1 Standard (final cut / premium quality). Stay on Veo unless the user explicitly asks for another model; seedance-2.0 for >8s narrative, kling-v3-pro for specific looks.",
enum: [
'veo-3.1',
'veo-3.1-fast',
'veo-3.1-lite',
'seedance-2.0',
'seedance-2.0-fast',
'kling-v3-pro',
'minimax-hailuo-2.3-pro',
'wan-2.2-a14b-turbo',
'ltx-2.3',
],
},
negativePrompt: {
type: 'string',
description:
'Things to exclude from the video/audio (Veo models), e.g. "no background music" to keep dialogue but drop Veo\'s invented music before overlaying your own track.',
},
outputs: {
type: 'object',
description:
'Workspace files to create or overwrite from returned code results or sandbox-created files.',
properties: {
files: {
type: 'array',
description: 'File outputs. Parent folders must already exist for create mode.',
items: {
type: 'object',
properties: {
format: {
type: 'string',
description: 'Optional serialization format for returned values.',
enum: ['json', 'csv', 'txt', 'md', 'html'],
},
mimeType: {
type: 'string',
description: 'Optional MIME type override when inference is not enough.',
},
mode: {
type: 'string',
description: 'Create a new file or overwrite an existing file at path.',
enum: ['create', 'overwrite'],
},
path: {
type: 'string',
description: 'Canonical destination VFS path, e.g. "files/Reports/chart.png".',
},
sandboxPath: {
type: 'string',
description:
'Optional full path to a file created inside the sandbox. Omit to save the code return value.',
},
},
required: ['path', 'mode'],
},
},
},
},
prompt: {
type: 'string',
description:
'Detailed description of the video to generate (scene, action, camera movement, style).',
},
promptOptimizer: {
type: 'boolean',
description: 'Enable prompt optimization for MiniMax models (default true).',
},
resolution: {
type: 'string',
description: 'Video resolution (model-dependent), e.g. 720p or 1080p.',
enum: ['720p', '1080p', '4k'],
},
},
required: ['prompt'],
},
resultSchema: undefined,
},
get_block_outputs: {
parameters: {
type: 'object',
properties: {
blockIds: {
type: 'array',
description:
'Optional array of block UUIDs. If provided, returns outputs only for those blocks. If not provided, returns outputs for all blocks in the workflow.',
items: {
type: 'string',
},
},
workflowId: {
type: 'string',
description:
'Optional workflow ID. If not provided, uses the current workflow in context.',
},
},
},
resultSchema: undefined,
},
get_block_upstream_references: {
parameters: {
type: 'object',
properties: {
blockIds: {
type: 'array',
description:
'Required array of block UUIDs (minimum 1). Returns what each block can reference based on its position in the workflow graph.',
items: {
type: 'string',
},
},
workflowId: {
type: 'string',
description:
'Optional workflow ID. If not provided, uses the current workflow in context.',
},
},
required: ['blockIds'],
},
resultSchema: undefined,
},
get_deployed_workflow_state: {
parameters: {
type: 'object',
properties: {
workflowId: {
type: 'string',
description:
'Optional workflow ID. If not provided, uses the current workflow in context.',
},
},
},
resultSchema: undefined,
},
get_deployment_log: {
parameters: {
type: 'object',
properties: {
workflowId: {
type: 'string',
description:
'Optional workflow ID. If not provided, uses the current workflow in context.',
},
},
},
resultSchema: undefined,
},
get_page_contents: {
parameters: {
type: 'object',
properties: {
include_highlights: {
type: 'boolean',
description: 'Include key highlights (default false)',
},
include_summary: {
type: 'boolean',
description: 'Include AI-generated summary (default false)',
},
include_text: {
type: 'boolean',
description: 'Include full page text (default true)',
},
urls: {
type: 'array',
description: 'URLs to get content from (max 10)',
items: {
type: 'string',
},
},
},
required: ['urls'],
},
resultSchema: undefined,
},
get_platform_actions: {
parameters: {
type: 'object',
properties: {},
},
resultSchema: undefined,
},
get_scheduled_task_logs: {
parameters: {
type: 'object',
properties: {
executionId: {
type: 'string',
description: 'Optional execution ID for a specific run.',
},
includeDetails: {
type: 'boolean',
description: 'Include tool calls, outputs, and cost details.',
},
jobId: {
type: 'string',
description: 'The scheduled task (schedule) ID to get logs for.',
},
limit: {
type: 'number',
description: 'Max number of entries (default: 3, max: 5)',
},
},
required: ['jobId'],
},
resultSchema: undefined,
},
get_workflow_data: {
parameters: {
type: 'object',
properties: {
data_type: {
type: 'string',
description: 'The type of workflow data to retrieve',
enum: ['global_variables', 'custom_tools', 'mcp_tools', 'files'],
},
workflowId: {
type: 'string',
description:
'Optional workflow ID. If not provided, uses the current workflow in context.',
},
},
required: ['data_type'],
},
resultSchema: undefined,
},
get_workflow_run_options: {
parameters: {
type: 'object',
properties: {
workflowId: {
type: 'string',
description:
'Optional workflow ID. If not provided, uses the current workflow in context.',
},
},
},
resultSchema: undefined,
},
glob: {
parameters: {
type: 'object',
properties: {
pattern: {
type: 'string',
description:
'Glob pattern to match file paths. Supports * (any segment) and ** (any depth).',
},
toolTitle: {
type: 'string',
description:
'Optional target-only UI phrase for the search row. The UI verb is supplied for you, so pass text like "workflow configs" or "knowledge bases", not a full sentence like "Finding workflow configs".',
},
},
required: ['pattern', 'toolTitle'],
},
resultSchema: undefined,
},
grep: {
parameters: {
type: 'object',
properties: {
context: {
type: 'number',
description:
"Number of lines to show before and after each match. Only applies to output_mode 'content'.",
},
ignoreCase: {
type: 'boolean',
description: 'Case insensitive search (default false).',
},
lineNumbers: {
type: 'boolean',
description:
"Include line numbers in output (default true). Only applies to output_mode 'content'.",
},
maxResults: {
type: 'number',
description: 'Maximum number of matches to return (default 50).',
},
output_mode: {
type: 'string',
description:
"Output mode: 'content' shows matching lines (default), 'files_with_matches' shows only file paths, 'count' shows match counts per file.",
enum: ['content', 'files_with_matches', 'count'],
},
path: {
type: 'string',
description:
"Optional scope. A prefix (e.g. 'workflows/', 'environment/', 'internal/') searches the VFS map under it. An exact single-file path under files/ or uploads/ (optionally with /content) searches that file's content only; folders and multi-file trees are rejected for content search.",
},
pattern: {
type: 'string',
description:
"Regex pattern to search for. Searches VFS map entries (workflow JSON, metadata, plans, memories) by default; searches a single file's extracted text when path is one files/ or uploads/ file leaf.",
},
toolTitle: {
type: 'string',
description:
'Optional target-only UI phrase for the search row. The UI verb is supplied for you, so pass text like "Slack integrations" or "deployed workflows", not a full sentence like "Searching for Slack integrations".',
},
},
required: ['pattern', 'toolTitle'],
},
resultSchema: undefined,
},
knowledge: {
parameters: {
properties: {
request: {
description: 'What knowledge base action is needed.',
type: 'string',
},
},
required: ['request'],
type: 'object',
},
resultSchema: undefined,
},
knowledge_base: {
parameters: {
type: 'object',
properties: {
args: {
type: 'object',
description: 'Arguments for the operation',
properties: {
apiKey: {
type: 'string',
description:
'API key for API-key-based connectors (required when connector auth mode is apiKey)',
},
chunkingConfig: {
type: 'object',
description: "Chunking configuration (optional for 'create')",
properties: {
maxSize: {
type: 'number',
description: 'Maximum chunk size (100-4000, default: 1024)',
default: 1024,
},
minSize: {
type: 'number',
description: 'Minimum chunk size (1-2000, default: 1)',
default: 1,
},
overlap: {
type: 'number',
description: 'Overlap between chunks (0-500, default: 200)',
default: 200,
},
},
},
connectorId: {
type: 'string',
description:
'Connector ID (required for update_connector, delete_connector, sync_connector)',
},
connectorStatus: {
type: 'string',
description: 'Connector status (optional for update_connector)',
enum: ['active', 'paused'],
},
connectorType: {
type: 'string',
description:
"Connector type from registry, e.g. 'confluence', 'google_drive', 'notion' (required for add_connector). Read knowledgebases/connectors/{type}.json for the config schema.",
},
credentialId: {
type: 'string',
description:
'OAuth credential ID from environment/credentials.json (required for OAuth connectors)',
},
description: {
type: 'string',
description: "Description of the knowledge base (optional for 'create')",
},
disabledTagIds: {
type: 'array',
description:
'Tag definition IDs to opt out of (optional for add_connector). See tagDefinitions in the connector schema.',
},
documentId: {
type: 'string',
description: 'Document ID (required for update_document)',
},
documentIds: {
type: 'array',
description: 'Document IDs (for batch delete_document)',
items: {
type: 'string',
},
},
enabled: {
type: 'boolean',
description: 'Enable/disable a document (optional for update_document)',
},
filePaths: {
type: 'array',
description:
'Canonical workspace file VFS paths to add as documents (for add_file), e.g. ["files/Docs/handbook.pdf"].',
items: {
type: 'string',
},
},
filename: {
type: 'string',
description: 'New filename for a document (optional for update_document)',
},
knowledgeBaseId: {
type: 'string',
description:
'Knowledge base ID (required for get, query, add_file, list_tags, create_tag, get_tag_usage)',
},
knowledgeBaseIds: {
type: 'array',
description: 'Knowledge base IDs (for batch delete)',
items: {
type: 'string',
},
},
name: {
type: 'string',
description: "Name of the knowledge base (required for 'create')",
},
query: {
type: 'string',
description: "Search query text (required for 'query')",
},
sourceConfig: {
type: 'object',
description:
'Connector-specific configuration matching the configFields in knowledgebases/connectors/{type}.json',
},
syncIntervalMinutes: {
type: 'number',
description:
'Sync interval in minutes. Accepted values: 60 (hourly), 360 (6h), 1440 (daily), 10080 (weekly), 0 (manual only). Default: 1440',
default: 1440,
},
tagDefinitionId: {
type: 'string',
description: 'Tag definition ID (required for update_tag, delete_tag)',
},
tagDisplayName: {
type: 'string',
description:
'Display name for the tag (required for create_tag, optional for update_tag)',
},
tagFieldType: {
type: 'string',
description:
'Field type: text, number, date, boolean (optional for create_tag, defaults to text)',
enum: ['text', 'number', 'date', 'boolean'],
},
topK: {
type: 'number',
description: 'Number of results to return (1-50, default: 5)',
default: 5,
},
workspaceId: {
type: 'string',
description:
"Workspace ID. Required for 'create' when there is no workspace in context; otherwise the current workspace context is used.",
},
},
},
operation: {
type: 'string',
description: 'The operation to perform',
enum: [
'create',
'get',
'query',
'add_file',
'update',
'delete',
'delete_document',
'update_document',
'list_tags',
'create_tag',
'update_tag',
'delete_tag',
'get_tag_usage',
'add_connector',
'update_connector',
'delete_connector',
'sync_connector',
],
},
},
required: ['operation'],
},
resultSchema: {
type: 'object',
properties: {
data: {
type: 'object',
description: 'Operation-specific result payload.',
},
message: {
type: 'string',
description: 'Human-readable outcome summary.',
},
success: {
type: 'boolean',
description: 'Whether the operation succeeded.',
},
},
required: ['success', 'message'],
},
},
list_file_folders: {
parameters: {
type: 'object',
properties: {
workspaceId: {
type: 'string',
description: 'Optional workspace ID. Defaults to the current workspace.',
},
},
},
resultSchema: undefined,
},
list_integration_tools: {
parameters: {
properties: {
integration: {
description:
'The integration service name — the folder under components/integrations/ (e.g. "slack", "gmail", "google_sheets"). Returns every operation\'s id, name, and description for that service.',
type: 'string',
},
},
required: ['integration'],
type: 'object',
},
resultSchema: undefined,
},
list_user_workspaces: {
parameters: {
type: 'object',
properties: {},
},
resultSchema: undefined,
},
list_workspace_mcp_servers: {
parameters: {
type: 'object',
properties: {
workspaceId: {
type: 'string',
description:
'Workspace ID. Required when no current workspace context is available, such as headless MCP calls.',
},
},
},
resultSchema: undefined,
},
load_deployment: {
parameters: {
type: 'object',
properties: {
version: {
type: 'string',
description:
'A string: a deployment version number (e.g. "5"), or "live" for the active deployment. (Unlike promote_to_live, which takes a numeric version, "live" is accepted here.)',
},
workflowId: {
type: 'string',
description:
'Optional workflow ID. If not provided, uses the current workflow in context.',
},
},
required: ['version'],
},
resultSchema: undefined,
},
load_integration_tool: {
parameters: {
properties: {
tool_ids: {
description:
'Exact integration tool ids to load before calling them, e.g. ["gmail_send_v2"]. Copy the "id" field verbatim from components/integrations/{service}/{operation}.json (including any version suffix).',
items: {
type: 'string',
},
type: 'array',
},
},
required: ['tool_ids'],
type: 'object',
},
resultSchema: undefined,
},
manage_credential: {
parameters: {
type: 'object',
properties: {
credentialId: {
type: 'string',
description: 'The credential ID (required for rename)',
},
credentialIds: {
type: 'array',
description: 'Array of credential IDs (for batch delete)',
items: {
type: 'string',
},
},
displayName: {
type: 'string',
description: 'New display name (required for rename)',
},
operation: {
type: 'string',
description: 'The operation to perform',
enum: ['rename', 'delete'],
},
},
required: ['operation'],
},
resultSchema: undefined,
},
manage_custom_tool: {
parameters: {
type: 'object',
properties: {
code: {
type: 'string',
description:
'The JavaScript code that executes when the tool is called (required for add). Parameters from schema are available as variables. Function body only - no signature or wrapping braces.',
},
operation: {
type: 'string',
description:
"The operation to perform: 'add', 'edit', 'list', or 'delete'. These verbs are tool-specific — manage_scheduled_task uses create/update instead of add/edit.",
enum: ['add', 'edit', 'delete', 'list'],
},
schema: {
type: 'object',
description: 'The tool schema in OpenAI function calling format (required for add).',
properties: {
function: {
type: 'object',
description: 'The function definition',
properties: {
description: {
type: 'string',
description: 'What the function does',
},
name: {
type: 'string',
description: 'The function name (camelCase)',
},
parameters: {
type: 'object',
description: 'The function parameters schema',
properties: {
properties: {
type: 'object',
description: 'Parameter definitions as key-value pairs',
},
required: {
type: 'array',
description: 'Array of required parameter names',
items: {
type: 'string',
},
},
type: {
type: 'string',
description: "Must be 'object'",
},
},
required: ['type', 'properties'],
},
},
required: ['name', 'parameters'],
},
type: {
type: 'string',
description: "Must be 'function'",
},
},
required: ['type', 'function'],
},
toolId: {
type: 'string',
description:
"The ID of the custom tool. Get it from the `list` operation or the `id` field inside the tool's VFS file (agent/custom-tools/{name}.json — the filename is the display name, not the id); get_workflow_data also returns it where that tool is available. Do not guess or construct it. Required for edit and delete; omit for add and list.",
},
toolIds: {
type: 'array',
description: 'Array of custom tool IDs (for batch delete)',
items: {
type: 'string',
},
},
},
required: ['operation'],
},
resultSchema: undefined,
},
manage_folder: {
parameters: {
type: 'object',
properties: {
destinationPath: {
type: 'string',
description:
'Destination parent folder\'s VFS path for move/create. Omit (or pass "workflows") to target the workspace root.',
},
folderId: {
type: 'string',
description:
'Target folder ID, used as a fallback when path is not given. Readable from a contained workflow\'s meta.json "folderId".',
},
name: {
type: 'string',
description:
'Folder name. Required for rename (the new name); for create when you pass a destination parent instead of a full path.',
},
operation: {
type: 'string',
description: 'The operation to perform.',
enum: ['create', 'rename', 'move', 'delete'],
},
parentId: {
type: 'string',
description:
'Destination parent folder ID, used as a fallback when destinationPath is not given.',
},
path: {
type: 'string',
description:
'Target folder\'s VFS path (e.g. "workflows/Marketing/Q3 Campaigns"), per-segment percent-encoded like every VFS path. Identifies the folder for rename/move/delete; for create it is the new folder\'s full path (its parent must already exist).',
},
},
required: ['operation'],
},
resultSchema: undefined,
},
manage_mcp_tool: {
parameters: {
type: 'object',
properties: {
config: {
type: 'object',
description: 'Required for add and edit. The MCP server configuration.',
properties: {
enabled: {
type: 'boolean',
description: 'Whether the server is enabled (default: true)',
},
headers: {
type: 'object',
description: 'Optional HTTP headers to send with requests (key-value pairs)',
},
name: {
type: 'string',
description: 'Display name for the MCP server',
},
timeout: {
type: 'number',
description: 'Request timeout in milliseconds (default: 30000)',
},
transport: {
type: 'string',
description: "Transport protocol: 'streamable-http' or 'sse'",
enum: ['streamable-http', 'sse'],
default: 'streamable-http',
},
url: {
type: 'string',
description: 'The MCP server endpoint URL (required for add)',
},
},
},
operation: {
type: 'string',
description:
"The operation to perform: 'add', 'edit', 'list', or 'delete'. These verbs are tool-specific — manage_scheduled_task uses create/update instead of add/edit.",
enum: ['add', 'edit', 'delete', 'list'],
},
serverId: {
type: 'string',
description:
"The MCP server's id — the `id` field inside the VFS file agent/mcp-servers/{name}.json (the {name} filename is the display name, not the id). Required for edit and delete; omit for add and list.",
},
},
required: ['operation'],
},
resultSchema: undefined,
},
manage_scheduled_task: {
parameters: {
type: 'object',
properties: {
args: {
type: 'object',
description:
'Operation-specific arguments. For create: {title, prompt, cron?, time?, timezone?, lifecycle?, successCondition?, maxRuns?}. For get/delete: {jobId}. For update: {jobId, title?, prompt?, cron?, timezone?, status?, lifecycle?, successCondition?, maxRuns?}. For list: no args needed.',
properties: {
cron: {
type: 'string',
description:
"Cron expression for a recurring scheduled task (e.g. '0 9 * * *'). Set exactly one of cron or time: recurring -> cron; one-time -> time.",
},
jobId: {
type: 'string',
description: 'Scheduled task ID (required for get, update)',
},
jobIds: {
type: 'array',
description: 'Array of scheduled task IDs (for batch delete)',
items: {
type: 'string',
},
},
lifecycle: {
type: 'string',
description:
"'persistent' (default) or 'until_complete'. Until_complete scheduled tasks stop when complete_scheduled_task is called.",
enum: ['persistent', 'until_complete'],
},
maxRuns: {
type: 'integer',
description: 'Max executions before auto-completing. Safety limit.',
},
prompt: {
type: 'string',
description: 'The prompt to execute when the scheduled task fires',
},
status: {
type: 'string',
description: 'Scheduled task status: active, paused',
enum: ['active', 'paused'],
},
successCondition: {
type: 'string',
description:
'What must happen for the scheduled task to be considered complete (until_complete lifecycle).',
},
time: {
type: 'string',
description:
"ISO 8601 datetime. One-time scheduled task -> set time and omit cron. May also anchor a recurring cron task's first-fire time.",
},
timezone: {
type: 'string',
description: 'IANA timezone (e.g. America/New_York). Defaults to UTC.',
},
title: {
type: 'string',
description: "Short descriptive title for the scheduled task (e.g. 'Email Poller')",
},
},
},
operation: {
type: 'string',
description:
'The operation to perform: create, list, get, update, delete. These verbs are tool-specific — the custom-tool/MCP/skill managers use add/edit instead of create/update.',
enum: ['create', 'list', 'get', 'update', 'delete'],
},
},
required: ['operation'],
},
resultSchema: undefined,
},
manage_skill: {
parameters: {
type: 'object',
properties: {
content: {
type: 'string',
description: 'Markdown instructions for the skill. Required for add, optional for edit.',
},
description: {
type: 'string',
description: 'Short description of the skill. Required for add, optional for edit.',
},
name: {
type: 'string',
description:
"Skill name in kebab-case (e.g. 'my-skill'). Required for add, optional for edit.",
},
operation: {
type: 'string',
description:
"The operation to perform: 'add', 'edit', 'list', or 'delete'. These verbs are tool-specific — manage_scheduled_task uses create/update instead of add/edit.",
enum: ['add', 'edit', 'delete', 'list'],
},
skillId: {
type: 'string',
description:
"The skill's id — the `id` field inside the VFS file agent/skills/{name}.json (the {name} filename is the display name, not the id). Required for edit and delete; omit for add and list.",
},
},
required: ['operation'],
},
resultSchema: undefined,
},
materialize_file: {
parameters: {
type: 'object',
properties: {
fileNames: {
type: 'array',
description:
'The names of the uploaded files to materialize (e.g. ["report.pdf", "data.csv"])',
items: {
type: 'string',
},
},
operation: {
type: 'string',
description:
'What to do with the file. "save" promotes it to a permanent files/ path. "import" imports a workflow JSON as a workspace workflow. Defaults to "save".',
enum: ['save', 'import'],
default: 'save',
},
},
required: ['fileNames'],
},
resultSchema: undefined,
},
media: {
parameters: {
properties: {
prompt: {
description:
"Optional brief instruction (one short sentence) to scope the task. The agent inherits the full conversation history — do NOT restate or rewrite conversation content, only add scoping the history doesn't convey.",
type: 'string',
},
},
type: 'object',
},
resultSchema: undefined,
},
move_file: {
parameters: {
type: 'object',
properties: {
destinationPath: {
type: 'string',
description:
'Canonical target folder path, e.g. "files/Images". Omit or pass "files" for root.',
},
paths: {
type: 'array',
description: 'Canonical workspace file VFS paths to move, e.g. ["files/photo.png"].',
items: {
type: 'string',
},
},
},
required: ['paths'],
},
resultSchema: undefined,
},
move_file_folder: {
parameters: {
type: 'object',
properties: {
destinationPath: {
type: 'string',
description:
'Canonical target parent folder path, e.g. "files/Archive". Omit or pass "files" for root.',
},
path: {
type: 'string',
description: 'Canonical folder VFS path to move, e.g. "files/Reports/2026".',
},
},
required: ['path'],
},
resultSchema: undefined,
},
move_workflow: {
parameters: {
type: 'object',
properties: {
folderId: {
type: 'string',
description: 'Target folder ID. Omit or pass empty string to move to workspace root.',
},
workflowIds: {
type: 'array',
description: 'The workflow IDs to move.',
items: {
type: 'string',
},
},
},
required: ['workflowIds'],
},
resultSchema: undefined,
},
oauth_get_auth_link: {
parameters: {
type: 'object',
properties: {
providerName: {
type: 'string',
description:
"The OAuth provider to connect. Pass the integration's provider value (e.g. `google-email`, `slack`); the service display name or providerId resolves case-insensitively/fuzzily, so avoid bare base providers like `google`.",
},
},
required: ['providerName'],
},
resultSchema: undefined,
},
oauth_request_access: {
parameters: {
type: 'object',
properties: {
providerName: {
type: 'string',
description:
"The OAuth provider to connect. Pass the integration's provider value (e.g. `google-email`, `slack`); the service display name or providerId resolves case-insensitively/fuzzily, so avoid bare base providers like `google`.",
},
},
required: ['providerName'],
},
resultSchema: undefined,
},
open_resource: {
parameters: {
type: 'object',
properties: {
resources: {
type: 'array',
description:
'Array of resources to open. Each item must have type and either id or, for files, path.',
items: {
type: 'object',
properties: {
id: {
type: 'string',
description: 'Canonical resource ID for non-file resources.',
},
path: {
type: 'string',
description:
'Encoded VFS path for type "file" (percent-encoded per segment, e.g. "files/Reports/Q4%20Report.pdf"). Copy it verbatim from glob/read/workspace context output — do not decode it to a display name or re-encode it.',
},
type: {
type: 'string',
description: 'The resource type.',
enum: ['workflow', 'table', 'knowledgebase', 'file', 'log', 'scheduledtask'],
},
},
required: ['type'],
},
},
},
required: ['resources'],
},
resultSchema: undefined,
},
promote_to_live: {
parameters: {
type: 'object',
properties: {
version: {
type: 'number',
description:
'The numeric deployment version number to promote to live (e.g. 5). "live" is not accepted here — pass the version number (use load_deployment to change the draft).',
},
workflowId: {
type: 'string',
description:
'Optional workflow ID. If not provided, uses the current workflow in context.',
},
},
required: ['version'],
},
resultSchema: undefined,
},
query_logs: {
parameters: {
type: 'object',
properties: {
blockId: {
type: 'string',
description: "Optional (view='full'): only return this block's span subtree.",
},
blockName: {
type: 'string',
description: "Optional (view='full'): only return spans for this block name.",
},
costOperator: {
type: 'string',
description: "Filter (view='list'): comparison operator for cost.",
enum: ['=', '>', '<', '>=', '<=', '!='],
},
costValue: {
type: 'number',
description: "Filter (view='list'): cost threshold paired with costOperator.",
},
cursor: {
type: 'string',
description: "Pagination cursor (view='list') from a prior response's nextCursor.",
},
durationOperator: {
type: 'string',
description: "Filter (view='list'): comparison operator for duration (ms).",
enum: ['=', '>', '<', '>=', '<=', '!='],
},
durationValue: {
type: 'number',
description:
"Filter (view='list'): duration threshold (ms) paired with durationOperator.",
},
endDate: {
type: 'string',
description: "Filter (view='list'): ISO end of the time range.",
},
executionId: {
type: 'string',
description:
"Required for 'overview'/'full': the execution to read. For 'list', an optional exact-match filter.",
},
folderIds: {
type: 'string',
description: "Filter (view='list'): comma-separated folder IDs (descendants included).",
},
folderName: {
type: 'string',
description: "Filter (view='list'): substring match on folder name.",
},
level: {
type: 'string',
description:
"Filter (view='list'): comma-separated levels: error, info, running, pending. Default all.",
},
limit: {
type: 'number',
description: "Max results (view='list'), 1-200 (default 100).",
},
pattern: {
type: 'string',
description:
"Optional separate parameter (not a 'view' value): with view 'overview' or 'full', greps the execution's trace spans (requires executionId), returning matching spans with snippets instead of the full log.",
},
search: {
type: 'string',
description: "Filter (view='list'): substring match on executionId.",
},
sortBy: {
type: 'string',
description: "Sort field (view='list').",
enum: ['date', 'duration', 'cost', 'status'],
},
sortOrder: {
type: 'string',
description: "Sort order (view='list').",
enum: ['asc', 'desc'],
},
startDate: {
type: 'string',
description: "Filter (view='list'): ISO start of the time range.",
},
triggers: {
type: 'string',
description: "Filter (view='list'): comma-separated trigger types.",
},
view: {
type: 'string',
description:
"Disclosure level: 'list' (summaries), 'overview' (one execution's trace tree, no I/O), or 'full' (one execution's trace spans with I/O).",
enum: ['list', 'overview', 'full'],
},
workflowIds: {
type: 'string',
description: "Filter (view='list'): comma-separated workflow IDs.",
},
workflowName: {
type: 'string',
description: "Filter (view='list'): substring match on workflow name.",
},
workspaceId: {
type: 'string',
description: 'Workspace ID to scope to.',
},
},
required: ['view'],
},
resultSchema: undefined,
},
read: {
parameters: {
type: 'object',
properties: {
limit: {
type: 'number',
description: 'Maximum number of lines to read.',
},
offset: {
type: 'number',
description: 'Line offset to start reading from (0-indexed).',
},
outputTable: {
type: 'string',
description:
'Table ID to import the file contents into (CSV/JSON). All existing rows are replaced. Example: "tbl_abc123"',
},
path: {
type: 'string',
description:
"Path to the VFS resource to read (e.g. 'workflows/My%20Workflow/state.json', 'files/Q4%20Report.pdf/content' for file bytes/parsed text, or 'uploads/data.csv' for a chat upload). Copy paths verbatim from glob/grep/read output.",
},
},
required: ['path'],
},
resultSchema: undefined,
},
redeploy: {
parameters: {
type: 'object',
properties: {
versionDescription: {
type: 'string',
description:
'REQUIRED: a concise (1-3 sentence) description of what changed in this deployment version. If unsure what changed, call diff_workflows(ref1: "live", ref2: "draft") first.',
},
versionName: {
type: 'string',
description:
'REQUIRED: a short human-readable name/label for this deployment version, shown in deployment history.',
},
workflowId: {
type: 'string',
description: 'Workflow ID to redeploy (required in workspace context)',
},
},
required: ['versionDescription', 'versionName'],
},
resultSchema: {
type: 'object',
properties: {
apiEndpoint: {
type: 'string',
description: 'Canonical workflow execution endpoint.',
},
baseUrl: {
type: 'string',
description: 'Base URL used to construct deployment URLs.',
},
deployedAt: {
type: 'string',
description: 'Deployment timestamp when the workflow is deployed.',
},
deploymentConfig: {
type: 'object',
description:
'Structured deployment configuration keyed by surface name. For API deploys this includes endpoint, auth, and sync/stream/async mode details.',
},
deploymentStatus: {
type: 'object',
description:
'Structured per-surface deployment status keyed by surface name, such as api.',
},
deploymentType: {
type: 'string',
description:
'Deployment surface this result describes. For deploy_api and redeploy this is always "api".',
},
examples: {
type: 'object',
description:
'Invocation examples keyed by surface name. For API deploys this includes curl examples for sync, stream, async, and polling.',
},
isDeployed: {
type: 'boolean',
description: 'Whether the workflow API is currently deployed after this tool call.',
},
version: {
type: 'number',
description: 'Deployment version for the current API deployment.',
},
workflowId: {
type: 'string',
description: 'Workflow ID that was deployed or undeployed.',
},
},
required: [
'workflowId',
'isDeployed',
'deploymentType',
'deploymentStatus',
'deploymentConfig',
'examples',
],
},
},
rename_file: {
parameters: {
type: 'object',
properties: {
newName: {
type: 'string',
description:
'New filename including extension, e.g. "draft_v2.md". Use move_file to move files between folders.',
},
path: {
type: 'string',
description:
'Canonical workspace file VFS path to rename, e.g. "files/Reports/draft.md".',
},
},
required: ['path', 'newName'],
},
resultSchema: {
type: 'object',
properties: {
data: {
type: 'object',
description: 'Contains id and the new name.',
},
message: {
type: 'string',
description: 'Human-readable outcome.',
},
success: {
type: 'boolean',
description: 'Whether the rename succeeded.',
},
},
required: ['success', 'message'],
},
},
rename_file_folder: {
parameters: {
type: 'object',
properties: {
name: {
type: 'string',
description: 'New folder name.',
},
path: {
type: 'string',
description: 'Canonical folder VFS path to rename, e.g. "files/Reports/Old".',
},
},
required: ['path', 'name'],
},
resultSchema: undefined,
},
rename_workflow: {
parameters: {
type: 'object',
properties: {
name: {
type: 'string',
description: 'The new name for the workflow.',
},
workflowId: {
type: 'string',
description: 'The workflow ID to rename.',
},
},
required: ['workflowId', 'name'],
},
resultSchema: undefined,
},
research: {
parameters: {
properties: {
topic: {
description: 'The topic to research.',
type: 'string',
},
},
required: ['topic'],
type: 'object',
},
resultSchema: undefined,
},
respond: {
parameters: {
additionalProperties: true,
properties: {
output: {
description:
'The result — facts, status, VFS paths to persisted data, whatever the caller needs to act on.',
type: 'string',
},
success: {
description: 'Whether the task completed successfully',
type: 'boolean',
},
type: {
description: 'Optional logical result type override',
type: 'string',
},
},
required: ['output', 'success'],
type: 'object',
},
resultSchema: undefined,
},
restore_resource: {
parameters: {
type: 'object',
properties: {
id: {
type: 'string',
description: 'The canonical resource ID to restore.',
},
type: {
type: 'string',
description: 'The resource type to restore.',
enum: ['workflow', 'table', 'file', 'knowledgebase', 'folder', 'file_folder'],
},
},
required: ['type', 'id'],
},
resultSchema: undefined,
},
run: {
parameters: {
properties: {
context: {
description: 'Pre-gathered context: workflow state, block IDs, input requirements.',
type: 'string',
},
request: {
description: 'What to run or what logs to check.',
type: 'string',
},
},
required: ['request'],
type: 'object',
},
resultSchema: undefined,
},
run_block: {
parameters: {
type: 'object',
properties: {
blockId: {
type: 'string',
description: 'The block ID to run in isolation.',
},
executionId: {
type: 'string',
description:
'Optional execution ID to load the snapshot from. Uses latest execution if omitted.',
},
useDeployedState: {
type: 'boolean',
description:
'When true, runs the deployed version instead of the live draft. Default: false (draft).',
},
workflowId: {
type: 'string',
description:
'Optional workflow ID to run. If not provided, uses the current workflow in context.',
},
workflow_input: {
type: 'object',
description: 'JSON object with key-value mappings where each key is an input field name',
},
},
required: ['blockId'],
},
resultSchema: undefined,
},
run_from_block: {
parameters: {
type: 'object',
properties: {
executionId: {
type: 'string',
description:
'Optional execution ID to load the snapshot from. Uses latest execution if omitted.',
},
startBlockId: {
type: 'string',
description: 'The block ID to start execution from.',
},
useDeployedState: {
type: 'boolean',
description:
'When true, runs the deployed version instead of the live draft. Default: false (draft).',
},
workflowId: {
type: 'string',
description:
'Optional workflow ID to run. If not provided, uses the current workflow in context.',
},
workflow_input: {
type: 'object',
description: 'JSON object with key-value mappings where each key is an input field name',
},
},
required: ['startBlockId'],
},
resultSchema: undefined,
},
run_workflow: {
parameters: {
type: 'object',
properties: {
inputFromExecutionId: {
type: 'string',
description:
'Reuse the recorded input from a past execution of this workflow (from query_logs) instead of supplying workflow_input — handy for replaying a run without retyping inputs. The reused input is re-validated against the trigger. Mutually exclusive with workflow_input and useMockPayload.',
},
triggerBlockId: {
type: 'string',
description:
'Trigger block ID to run from (from get_workflow_run_options). Required when the workflow has multiple entrypoints.',
},
useDeployedState: {
type: 'boolean',
description:
'When true, runs the deployed version instead of the live draft. Default: false (draft).',
},
useMockPayload: {
type: 'boolean',
description:
"When true, run with the trigger's generated mock payload instead of workflow_input. Prefer building your own workflow_input; use this only when you can't.",
},
workflowId: {
type: 'string',
description:
'Optional workflow ID to run. If not provided, uses the current workflow in context.',
},
workflow_input: {
type: 'object',
description:
"JSON object matching the target trigger's inputSchema (from get_workflow_run_options). For external/webhook triggers this is the event payload; for API/Input triggers it is the form fields.",
},
},
},
resultSchema: undefined,
},
run_workflow_until_block: {
parameters: {
type: 'object',
properties: {
inputFromExecutionId: {
type: 'string',
description:
'Reuse the recorded input from a past execution of this workflow (from query_logs) instead of supplying workflow_input. The reused input is re-validated against the trigger. Mutually exclusive with workflow_input and useMockPayload.',
},
stopAfterBlockId: {
type: 'string',
description: 'The block ID to stop after. Execution halts once this block completes.',
},
triggerBlockId: {
type: 'string',
description:
'Trigger block ID to run from (from get_workflow_run_options). Required when the workflow has multiple entrypoints.',
},
useDeployedState: {
type: 'boolean',
description:
'When true, runs the deployed version instead of the live draft. Default: false (draft).',
},
useMockPayload: {
type: 'boolean',
description:
"When true, run with the trigger's generated mock payload instead of workflow_input. Prefer building your own workflow_input; use this only when you can't.",
},
workflowId: {
type: 'string',
description:
'Optional workflow ID to run. If not provided, uses the current workflow in context.',
},
workflow_input: {
type: 'object',
description:
"JSON object matching the target trigger's inputSchema (from get_workflow_run_options). For external/webhook triggers this is the event payload; for API/Input triggers it is the form fields.",
},
},
required: ['stopAfterBlockId'],
},
resultSchema: undefined,
},
scheduled_task: {
parameters: {
properties: {
request: {
description: 'What scheduled task action is needed.',
type: 'string',
},
},
required: ['request'],
type: 'object',
},
resultSchema: undefined,
},
scrape_page: {
parameters: {
type: 'object',
properties: {
include_links: {
type: 'boolean',
description: 'Extract all links from the page (default false)',
},
url: {
type: 'string',
description: 'The URL to scrape (must include https://)',
},
wait_for: {
type: 'string',
description: 'CSS selector to wait for before scraping (for JS-heavy pages)',
},
},
required: ['url'],
},
resultSchema: undefined,
},
search_documentation: {
parameters: {
type: 'object',
properties: {
query: {
type: 'string',
description: 'The search query',
},
topK: {
type: 'number',
description: 'Number of results (max 10)',
},
},
required: ['query'],
},
resultSchema: undefined,
},
search_library_docs: {
parameters: {
type: 'object',
properties: {
library_name: {
type: 'string',
description: "Name of the library to search for (e.g., 'nextjs', 'stripe', 'langchain')",
},
query: {
type: 'string',
description: 'The question or topic to find documentation for - be specific',
},
version: {
type: 'string',
description: "Specific version (optional, e.g., '14', 'v2')",
},
},
required: ['library_name', 'query'],
},
resultSchema: undefined,
},
search_online: {
parameters: {
type: 'object',
properties: {
category: {
type: 'string',
description: 'Filter by category',
enum: [
'news',
'tweet',
'github',
'company',
'research paper',
'linkedin profile',
'pdf',
'personal site',
],
},
include_text: {
type: 'boolean',
description: 'Include page text content (default true)',
},
num_results: {
type: 'number',
description: 'Number of results (default 10, max 25)',
},
query: {
type: 'string',
description: 'Natural language search query',
},
toolTitle: {
type: 'string',
description:
"Required short UI label fragment (e.g. 'Slack integrations'), not a full sentence.",
},
},
required: ['query', 'toolTitle'],
},
resultSchema: undefined,
},
search_patterns: {
parameters: {
type: 'object',
properties: {
limit: {
type: 'integer',
description: 'Maximum number of unique pattern examples to return (defaults to 3).',
},
queries: {
type: 'array',
description:
'Up to 3 descriptive strings explaining the workflow pattern(s) you need. Focus on intent and desired outcomes.',
items: {
type: 'string',
description: 'Example: "how to automate wealthbox meeting notes into follow-up tasks"',
},
},
},
required: ['queries'],
},
resultSchema: undefined,
},
set_block_enabled: {
parameters: {
type: 'object',
properties: {
blockId: {
type: 'string',
description: 'The block ID whose enabled state should be changed.',
},
enabled: {
type: 'boolean',
description: 'Set to true to enable the block, or false to disable it.',
},
workflowId: {
type: 'string',
description:
'Optional workflow ID to edit. If not provided, uses the current workflow in context.',
},
},
required: ['blockId', 'enabled'],
},
resultSchema: undefined,
},
set_environment_variables: {
parameters: {
type: 'object',
properties: {
scope: {
type: 'string',
description:
'Whether to set workspace or personal environment variables. Defaults to workspace.',
enum: ['personal', 'workspace'],
default: 'workspace',
},
variables: {
type: 'array',
description: 'List of env vars to set',
items: {
type: 'object',
properties: {
name: {
type: 'string',
description: 'Variable name',
},
value: {
type: 'string',
description: 'Variable value',
},
},
required: ['name', 'value'],
},
},
},
required: ['variables'],
},
resultSchema: undefined,
},
set_global_workflow_variables: {
parameters: {
type: 'object',
properties: {
operations: {
type: 'array',
description: 'List of operations to apply',
items: {
type: 'object',
properties: {
name: {
type: 'string',
description: 'Variable name.',
},
operation: {
type: 'string',
enum: ['add', 'delete', 'edit'],
},
type: {
type: 'string',
description: 'Variable type. Required for add/edit; ignored for delete.',
enum: ['plain', 'number', 'boolean', 'array', 'object'],
},
value: {
type: 'string',
description: 'Variable value. Required for add/edit; ignored for delete.',
},
},
required: ['operation', 'name'],
},
},
workflowId: {
type: 'string',
description:
'Optional workflow ID. If not provided, uses the current workflow in context.',
},
},
required: ['operations'],
},
resultSchema: undefined,
},
superagent: {
parameters: {
properties: {
task: {
description:
"A single sentence — the agent has full conversation context. Do NOT pre-read credentials or look up configs. Example: 'send the email we discussed' or 'check my calendar for tomorrow'.",
type: 'string',
},
},
required: ['task'],
type: 'object',
},
resultSchema: undefined,
},
table: {
parameters: {
properties: {
request: {
description: 'What table action is needed.',
type: 'string',
},
},
required: ['request'],
type: 'object',
},
resultSchema: undefined,
},
update_deployment_version: {
parameters: {
type: 'object',
properties: {
description: {
type: 'string',
description:
'New description for the deployment version. Provide name and/or description.',
},
name: {
type: 'string',
description:
'New name/label for the deployment version. Provide name and/or description.',
},
version: {
type: 'number',
description:
'The numeric deployment version number to update (use get_deployment_log to find it).',
},
workflowId: {
type: 'string',
description:
'Optional workflow ID. If not provided, uses the current workflow in context.',
},
},
required: ['version'],
},
resultSchema: undefined,
},
update_scheduled_task_history: {
parameters: {
type: 'object',
properties: {
jobId: {
type: 'string',
description: 'The scheduled task ID.',
},
summary: {
type: 'string',
description:
"A concise summary of what was done this run (e.g., 'Sent follow-up emails to 3 leads: Alice, Bob, Carol').",
},
},
required: ['jobId', 'summary'],
},
resultSchema: undefined,
},
update_workspace_mcp_server: {
parameters: {
type: 'object',
properties: {
description: {
type: 'string',
description: 'New description for the server',
},
isPublic: {
type: 'boolean',
description: 'Whether the server is publicly accessible',
},
name: {
type: 'string',
description: 'New name for the server',
},
serverId: {
type: 'string',
description: 'Required: the MCP server ID to update',
},
},
required: ['serverId'],
},
resultSchema: undefined,
},
user_memory: {
parameters: {
type: 'object',
properties: {
confidence: {
type: 'number',
description: 'Confidence level 0-1 (default 1.0 for explicit, 0.8 for inferred)',
},
correct_value: {
type: 'string',
description:
"The correct value to replace the wrong one (for 'correct' operation). Requires `key` (the memory to replace).",
},
key: {
type: 'string',
description: "Unique key for the memory (e.g., 'preferred_model', 'slack_credential')",
},
limit: {
type: 'number',
description: 'Number of results for search (default 10)',
},
memory_type: {
type: 'string',
description: "Type of memory: 'preference', 'entity', 'history', or 'correction'",
enum: ['preference', 'entity', 'history', 'correction'],
},
operation: {
type: 'string',
description: "Operation: 'add', 'search', 'delete', 'correct', or 'list'",
enum: ['add', 'search', 'delete', 'correct', 'list'],
},
query: {
type: 'string',
description: 'Search query to find relevant memories',
},
source: {
type: 'string',
description: "Source: 'explicit' (user told you) or 'inferred' (you observed)",
enum: ['explicit', 'inferred'],
},
value: {
type: 'string',
description: 'Value to remember',
},
},
required: ['operation'],
},
resultSchema: undefined,
},
user_table: {
parameters: {
type: 'object',
properties: {
args: {
type: 'object',
description: 'Arguments for the operation',
properties: {
autoRun: {
type: 'boolean',
description:
"Optional flag for add_workflow_group, add_enrichment, and update_workflow_group. On add (workflow group or enrichment): when true, existing rows whose dependencies are already filled run immediately; default false stages the group silently — call run_column when ready to fire rows. On update: toggle a group's auto-fire behavior on an existing group — false stages it (no auto-runs on dep satisfaction; only manual run_column fires rows), true re-enables auto-fire (rows whose deps fill will be scheduled). Set true on add only if the user explicitly asked to start runs immediately.",
},
blockId: {
type: 'string',
description:
'Source block ID inside the workflow. Used by add_workflow_group_output.',
},
column: {
type: 'object',
description: 'Column definition for add_column: { name, type, unique?, position? }',
},
columnName: {
type: 'string',
description:
'Column name. Required for rename_column, update_column, and delete_workflow_group_output (the bound column to drop). Optional for add_workflow_group_output (auto-derived from path when omitted). Use columnNames array for batch delete_column.',
},
columnNames: {
type: 'array',
description:
'Array of column names to delete at once (for delete_column). Preferred over columnName when deleting multiple columns.',
},
data: {
type: 'object',
description: 'Row data as key-value pairs (required for insert_row, update_row)',
},
dependencies: {
type: 'object',
description:
"Dependencies the group requires before running a row. { columns?: string[] } lists input column names that must be filled. Workflow output columns count too — depend on the column produced by an upstream group, not the group itself. The dep graph is column-induced. A group can't depend on its own output columns. Used by add_workflow_group and update_workflow_group, and optionally by add_enrichment (omit and the handler defaults deps to the mapped input columns).",
properties: {
columns: {
type: 'array',
description:
'Input column names that must be filled before the group runs. Plain columns and upstream-group output columns are both valid here.',
items: {
type: 'string',
},
},
},
},
description: {
type: 'string',
description: "Table description (optional for 'create')",
},
enrichmentId: {
type: 'string',
description:
"Enrichment registry ID for add_enrichment. Discover the available IDs (and each one's inputs/outputs) via list_enrichments first — don't hardcode. Examples: work-email, phone-number, company-domain, company-info.",
},
filePath: {
type: 'string',
description:
'Canonical workspace file VFS path for create_from_file/import_file, e.g. files/{path}/{name}.',
},
filter: {
type: 'object',
description:
'MongoDB-style filter for query_rows, update_rows_by_filter, delete_rows_by_filter',
},
groupId: {
type: 'string',
description:
'Workflow group ID. Required for update_workflow_group, delete_workflow_group, add_workflow_group_output, delete_workflow_group_output.',
},
groupIds: {
type: 'array',
description:
'Array of workflow group IDs. Required for run_column — non-empty list of columns to run.',
items: {
type: 'string',
},
},
inputMappings: {
type: 'array',
description:
'For add_enrichment: maps each enrichment input to an existing table column. Each item is { inputName, columnName } where inputName is the enrichment input id (from list_enrichments) and columnName is an existing column on the table. Provide a mapping for every required input. (The field is named inputName for consistency with workflow-group input mappings; for enrichments it holds the enrichment input id.)',
items: {
type: 'object',
properties: {
columnName: {
type: 'string',
description: 'Existing table column name that supplies this input.',
},
inputName: {
type: 'string',
description: 'Enrichment input id to bind (from list_enrichments).',
},
},
required: ['inputName', 'columnName'],
},
},
limit: {
type: 'number',
description:
'Maximum rows to return or affect (optional, default 100). Omit on update_rows_by_filter / delete_rows_by_filter to act on every match.',
},
mapping: {
type: 'object',
description:
'Optional explicit CSV-header → table-column mapping for import_file, as { "csvHeader": "columnName" | null }. A string maps the CSV header to that table column; null skips that CSV header (it won\'t be imported); omit a header entirely to fall back to auto-mapping by sanitized name (case-insensitive).',
additionalProperties: {
type: ['string', 'null'],
description:
"Target column name on the table. null skips that CSV header (it won't be imported); omit it entirely to fall back to auto-mapping.",
},
},
mappingUpdates: {
type: 'array',
description:
"Surgical per-output remap for update_workflow_group. Each entry repoints ONE existing output column to a new (blockId, path) without touching the rest of the group. Use this when the user wants to swap which block output flows into a column (e.g. 'point the score column at the new agent block') — the bound column stays, only its source pair changes. Stale row data for remapped columns is cleared and backfilled from saved execution logs where possible (no re-run needed). Use this INSTEAD of resending the full outputs array when the change is scoped to a few columns; use outputs only when the whole group's output set is being restructured. Discover valid (blockId, path) pairs via list_workflow_outputs first.",
items: {
type: 'object',
properties: {
blockId: {
type: 'string',
description: 'New source block ID for this column.',
},
columnName: {
type: 'string',
description:
'The existing output column to remap. Must already be bound to this group.',
},
path: {
type: 'string',
description: 'New dotted output path on the new block.',
},
},
required: ['columnName', 'blockId', 'path'],
},
},
mode: {
type: 'string',
description:
"Import mode for import_file. 'append' (default) adds rows; 'replace' truncates existing rows in a transaction before inserting the new rows.",
enum: ['append', 'replace'],
},
name: {
type: 'string',
description:
"Table name (required for 'create'). Also the optional display name for add_enrichment — defaults to the enrichment's registry name when omitted.",
},
newName: {
type: 'string',
description:
'New name. Required for rename_column (new column name) and for rename (new table name).',
},
newType: {
type: 'string',
description:
'New column type (optional for update_column). Types: string, number, boolean, date, json',
},
offset: {
type: 'number',
description: 'Number of rows to skip (optional for query_rows, default 0)',
},
outputColumnNames: {
type: 'object',
description:
'Optional output column name overrides for add_enrichment, as { "<outputId>": "<columnName>" }. Omit to use each enrichment output\'s default name.',
additionalProperties: {
type: 'string',
description: 'Target column name for this enrichment output id.',
},
},
outputFormat: {
type: 'string',
description:
'Explicit format override for outputPath. Usually unnecessary — the file extension determines the format automatically. Only use this to force a different format than what the extension implies.',
enum: ['json', 'csv', 'txt', 'md', 'html'],
},
outputPath: {
type: 'string',
description:
'Pipe query_rows results directly to a NEW workspace file. The format is auto-inferred from the file extension: .csv → CSV, .json → JSON, .md → Markdown, etc. Use a root output path like "files/export.csv" — nested output paths are not supported.',
},
outputs: {
type: 'array',
description:
"Outputs to surface as columns. Each entry maps a workflow block output to a table column: { blockId, path, columnName?, columnType? }. blockId is the source block; path is the dotted output path; columnName auto-derives from the path when omitted; columnType defaults from the leaf type when omitted. Used by add_workflow_group for the full output set. For update_workflow_group, prefer add_workflow_group_output / delete_workflow_group_output for individual outputs and mappingUpdates for surgical remap; only pass outputs here when restructuring the whole group's output set in one shot. If unsure about valid (blockId, path) pairs, call list_workflow_outputs first — paths are validated against the live workflow and invalid picks return an error with the valid options. For Agent blocks with structured outputs, the structured fields appear as top-level paths (e.g. summary, industry); there is NO response.content path on a structured agent.",
items: {
type: 'object',
properties: {
blockId: {
type: 'string',
description: 'Source block ID inside the workflow.',
},
columnName: {
type: 'string',
description:
'Optional target column name. Auto-derived from the path when omitted.',
},
columnType: {
type: 'string',
description: 'Optional column type. Defaults from the leaf type when omitted.',
enum: ['string', 'number', 'boolean', 'date', 'json'],
},
path: {
type: 'string',
description: 'Dotted output path on the block.',
},
},
required: ['blockId', 'path'],
},
},
path: {
type: 'string',
description: 'Dotted output path on the block. Used by add_workflow_group_output.',
},
position: {
type: 'integer',
description:
'Zero-based index at which to insert the row (optional, insert_row only). Rows at and below that index shift down. Omit to append at the end.',
},
positions: {
type: 'array',
description:
'Per-row insertion indices for batch_insert_rows (optional). Must be the same length as rows and contain no duplicates. Values are final positions in the resulting table — lower-index shifts are applied automatically. Omit to append all rows at the end.',
items: {
type: 'integer',
},
},
rowId: {
type: 'string',
description:
"Row ID. Required for get_row, update_row, delete_row, and for cancel_table_runs when scope:'row'.",
},
rowIds: {
type: 'array',
description:
'Array of row IDs. Used by batch_delete_rows (rows to delete) and run_column (optional row scope — when omitted, runs across the whole table; when provided, only these rows are candidates and the server eligibility predicate still applies).',
items: {
type: 'string',
},
},
rows: {
type: 'array',
description: 'Array of row data objects (required for batch_insert_rows)',
},
runMode: {
type: 'string',
description:
"Run mode for run_column. 'incomplete' (default) re-runs only rows that never produced output or last failed; 'all' re-runs every dep-satisfied row.",
enum: ['incomplete', 'all'],
},
schema: {
type: 'object',
description:
"Table schema with columns array (required for 'create'). Each column: { name, type, unique? }",
},
scope: {
type: 'string',
description:
"Cancellation scope for cancel_table_runs. 'all' cancels in-flight runs across the whole table; 'row' cancels only the row identified by rowId.",
enum: ['all', 'row'],
},
sort: {
type: 'object',
description:
"Sort specification as { field: 'asc' | 'desc' } (optional for query_rows)",
},
tableId: {
type: 'string',
description:
"Table ID (required for most operations except 'create' and batch 'delete')",
},
tableIds: {
type: 'array',
description: 'Array of table IDs (for batch delete)',
items: {
type: 'string',
},
},
unique: {
type: 'boolean',
description: 'Set column unique constraint (optional for update_column)',
},
updates: {
type: 'array',
description:
'Array of per-row updates: [{ rowId, data: { col: val } }] (for batch_update_rows)',
},
values: {
type: 'object',
description:
'Map of rowId to value for single-column batch update: { "rowId1": val1, "rowId2": val2 } (for batch_update_rows with columnName)',
},
workflowId: {
type: 'string',
description:
'ID of the workflow (required for add_workflow_group and list_workflow_outputs).',
},
},
},
operation: {
type: 'string',
description: 'The operation to perform',
enum: [
'create',
'create_from_file',
'import_file',
'get',
'get_schema',
'delete',
'rename',
'insert_row',
'batch_insert_rows',
'get_row',
'query_rows',
'update_row',
'delete_row',
'update_rows_by_filter',
'delete_rows_by_filter',
'batch_update_rows',
'batch_delete_rows',
'add_column',
'rename_column',
'delete_column',
'update_column',
'add_workflow_group',
'update_workflow_group',
'delete_workflow_group',
'add_workflow_group_output',
'delete_workflow_group_output',
'run_column',
'cancel_table_runs',
'list_workflow_outputs',
'list_enrichments',
'add_enrichment',
],
},
},
required: ['operation', 'args'],
},
resultSchema: {
type: 'object',
properties: {
data: {
type: 'object',
description: 'Operation-specific result payload.',
},
message: {
type: 'string',
description: 'Human-readable outcome summary.',
},
success: {
type: 'boolean',
description: 'Whether the operation succeeded.',
},
},
required: ['success', 'message'],
},
},
workflow: {
parameters: {
properties: {
prompt: {
description:
"Optional brief instruction (one short sentence) to scope the task. The agent inherits the full conversation history — do NOT restate or rewrite conversation content, only add scoping the history doesn't convey.",
type: 'string',
},
},
type: 'object',
},
resultSchema: undefined,
},
workspace_file: {
parameters: {
type: 'object',
properties: {
operation: {
type: 'string',
description: 'The file operation to perform.',
enum: ['append', 'update', 'patch'],
},
target: {
type: 'object',
description: 'Explicit file target. Use kind=path + path for existing files.',
properties: {
kind: {
type: 'string',
description: 'How the file target is identified.',
enum: ['path'],
},
path: {
type: 'string',
description:
'Canonical existing workspace file VFS path, e.g. "files/Reports/report.md". Required when target.kind=path.',
},
},
required: ['kind'],
},
title: {
type: 'string',
description:
'Required short UI label for this content unit, e.g. "Chapter 1", "Slide 3", or "Fix footer spacing".',
},
contentType: {
type: 'string',
description:
'Optional MIME type override. Usually omit and let the system infer from the target file extension.',
enum: [
'text/markdown',
'text/html',
'text/plain',
'application/json',
'text/csv',
'application/vnd.openxmlformats-officedocument.presentationml.presentation',
'application/vnd.openxmlformats-officedocument.wordprocessingml.document',
'application/pdf',
],
},
edit: {
type: 'object',
description:
'Patch metadata. Use strategy=search_replace for exact text replacement, or strategy=anchored for line-based inserts/replacements/deletions. The actual replacement/insert content is provided via the paired edit_content tool call.',
properties: {
after_anchor: {
type: 'string',
description:
'Boundary line kept after inserted replacement content. Required for mode=replace_between.',
},
anchor: {
type: 'string',
description:
'Anchor line after which new content is inserted. Required for mode=insert_after.',
},
before_anchor: {
type: 'string',
description:
'Boundary line kept before inserted replacement content. Required for mode=replace_between.',
},
end_anchor: {
type: 'string',
description: 'First line to keep after deletion. Required for mode=delete_between.',
},
mode: {
type: 'string',
description: 'Anchored edit mode when strategy=anchored.',
enum: ['replace_between', 'insert_after', 'delete_between'],
},
occurrence: {
type: 'number',
description: '1-based occurrence for repeated anchor lines. Optional; defaults to 1.',
},
replaceAll: {
type: 'boolean',
description:
'When true and strategy=search_replace, replace every match instead of requiring a unique single match.',
},
search: {
type: 'string',
description:
'Exact text to find when strategy=search_replace. Must match exactly once unless replaceAll=true.',
},
start_anchor: {
type: 'string',
description: 'First line to delete. Required for mode=delete_between.',
},
strategy: {
type: 'string',
description: 'Patch strategy.',
enum: ['search_replace', 'anchored'],
},
},
},
},
required: ['operation', 'target', 'title'],
},
resultSchema: {
type: 'object',
properties: {
data: {
type: 'object',
description:
'Optional operation metadata such as file id, file name, size, and content type.',
},
message: {
type: 'string',
description: 'Human-readable summary of the outcome.',
},
success: {
type: 'boolean',
description: 'Whether the file operation succeeded.',
},
},
required: ['success', 'message'],
},
},
}
Reference in New Issue View Git Blame Copy Permalink