Blinkbox

Execute JavaScript code and return the result. Useful for data transformation, calculations, string manipulation, JSON processing, and algorithm implementation. Code runs in a sandboxed Node.js environment with access to standard built-ins but no file system or network access.

Generate an image using DALL-E 3. Returns a URL to the generated image. Use this when the user asks you to create, draw, or generate an image or illustration. Requires an OpenAI credential ID.

Get full details of one automation by name or id: status, trigger, and its steps.

Check the status and result of a workflow run by its execution id.

Get the step-by-step logs for a workflow run by execution id.

Get everything needed to configure one node: its config fields (key, type, whether it's required, an example value, what it means), the output fields later steps can reference as {{ $json.field }}, whether it needs a credential and whether the user already has one, its output handles, and whether it exposes app actions. This is the node's configuration panel expressed as data — read it before filling a node's config instead of guessing field names.

Make an HTTP request to any URL or API endpoint. Supports GET, POST, PUT, PATCH, DELETE with custom headers and JSON body. Use this to call external APIs, fetch web pages, or send data to services.

List the user's Blinkbox automations (workflows) with their on/off status, trigger, and id.

List the user's saved credentials by name, type and id. Secrets are never returned. Pass a node key to see only the credentials that node can use, plus what's still missing for it.

List recent runs of an automation (by name or id) with their statuses.

List the operations an app node can perform — e.g. slack → send_message, list_channels — with what each one does, the OAuth scopes it needs, and which is the default. Also lists the resource pickers the node offers (channels, sheets, bases…) that resolve live from the user's credential. Use this to choose a node's 'operation' value and know which parameters go with it.

Browse or search Blinkbox's node catalog — the building blocks of a workflow. Only nodes the user can actually reach from a picker are listed, so anything returned here is safe to build with. Filter by picker ('trigger' = what starts a workflow, 'action' = a step, 'agent' = AI-agent models/memory/tools), by category, or by free-text search over the name and key. Use this to find real node keys before create_automation, then get_node to learn how to configure one.

Evaluate mathematical expressions safely. Supports arithmetic (+, -, *, /, %, ^), parentheses, and functions: sqrt, abs, ceil, floor, round, min, max, log, ln, sin, cos, tan, pi, e. Use this for any calculation — do NOT attempt mental math.

Retrieve a previously stored fact by its key. Use this to look up information you stored earlier with 'remember'.

Store a fact or piece of information for later use in this session. Call this whenever you learn something important that you'll need to reference again.

Rename an automation (look it up by current name or id).

Run an automation now by name or id (works whether it's active or not). Optionally pass input data. Returns the execution id.

Execute sandboxed JavaScript code in an isolated V8 environment. Input is available as $input, output via $output. No filesystem or network access.

Trigger another BlinkBox automation workflow. Use this when the user wants to kick off a sub-process, delegate to another workflow, or chain automations. Can only trigger workflows in the same workspace.

Send an email using the workspace's Gmail credential. Use this when the user asks you to email someone, send a message, or notify by email. Requires a Gmail OAuth credential ID.

Fetch any URL and return an AI-generated summary of its content. More efficient than web_search + reading results manually. Use this when you need to understand what a specific webpage contains.

Convert text to spoken audio using OpenAI TTS. Returns base64-encoded MP3 audio. Use this when the user asks you to narrate, read aloud, or create audio from text. Requires an OpenAI credential ID.

Use this to reason through a problem step by step before acting. Write your thoughts as the input — no external action is taken. The output is your own thought returned back. Use this to plan, reflect, or work through complex logic.

Scrape and extract content from a web page. Supports browser rendering (Puppeteer) and fast reader (Jina) engines. Returns page content as markdown, text, or HTML.

Search the internet for current, real-time information. Returns titles, URLs, content snippets, and an AI-generated summary answer. Use this when you need up-to-date facts, news, or information not in your training data.

Search Wikipedia for factual information about people, places, events, concepts, and history. Returns a concise summary from the Wikipedia article.

Read and write key-value data to workspace-scoped persistent memory (Redis). Use 'read' to retrieve previously stored values, 'write' to store new data, 'delete' to remove a key, or 'list' to see all stored keys. Data persists across agent executions within the same workspace.