Navigation

Mobile Docs

Catalog

Agent

Build an agent

Create powerful AI agents using any LLM provider with customizable system prompts and tool integrations.

Agentsagent

Usage

01Add the block to your workflow and connect it to the upstream step.
02Configure any required credentials or tokens in the inputs.
03Fill in required inputs and optional parameters for the run.
04Run a test execution, inspect outputs, and iterate before deploying.
05Deploy the agent block with monitoring enabled in production.

Inputs (UI)

Agent Profile

agent-profile-selector

Layout: full

System Prompt

long-input

Placeholder: You are a helpful AI assistant. Be concise and professional.

Layout: full

Condition: agentProfileId = ""

User Prompt

long-input

Placeholder: Enter user message here. Use %input% for variables.

Layout: full

Model

dropdown

Layout: full

Options: Dynamic options

Tools

tool-input

Layout: full

Web Research

switch

Enable autonomous web research capability

Layout: full

Research Sources

checkbox-list

Layout: half

Condition: webResearchEnabled = true

Options: Web Search, Web Scrape, Wikipedia, Reddit

Response Format (Optional)

code

Placeholder: Leave empty for natural language response, or define JSON schema: { "type": "object", "properties": { "answer": {"type": "string"} }, "additionalProperties": false }

Layout: full

Knowledge Sources

knowledge-source-input

Layout: full

Max Search Results

slider

Layout: half

Condition: knowledgeSources = ""

Similarity Threshold

slider

Layout: half

Condition: knowledgeSources = ""

short-input

Placeholder: customer-support, billing, technical

Layout: half

Primary response type:

{
  "content": "string",
  "model": "string",
  "tokens": "json",
  "toolCalls": "json"
}

Conditional output based on: responseFormat

When empty: {"content":"string","model":"string","tokens":"json","toolCalls":"json"} | When filled: "json"

Tool Access

openai_chatanthropic_chatgoogle_chatxai_chatdeepseek_chatdeepseek_reasoner