Agent Reference
Agents are the building blocks of every Ema workflow. Each agent performs a specific task -- searching knowledge bases, generating responses, classifying input, calling external tools, or validating results. You compose agents into workflows by connecting their inputs and outputs on the Generative Workflow Engine (GWE) canvas.
This reference catalogs every agent available on the Ema Builder Platform, organized by function.
Agent Categories
| Category | Description | Example Agents |
|---|---|---|
| Search & Retrieval | Find information from knowledge bases, the web, or connected applications | Knowledge Search, Agentic Search and Respond, Deep Web Search, Document Synthesis |
| Response Generation | Produce text responses grounded in context, search results, or templates | Respond to a Query, Respond using Search Results, Fixed Response, Custom Agent |
| Classification & Routing | Categorize inputs and route workflows based on content | Categorizer, Categorize Conversations and Route, Thread Categorizer |
| Data Processing | Transform, extract, and combine data between workflow steps | Convert to Text, Combine Text, Combine Search Results, JSON Extractor, Extract Entities, Custom Code Agent |
| Document Generation | Create formatted documents, charts, and reports | Generate Document, Chart Builder |
| External Integration | Execute actions in third-party applications | Intelligent Actions, App Navigator, Send Email, Sales Intelligence |
| Validation & QA | Check outputs against rules, quality criteria, or business logic | Response Validator, Rule Validation, Abstain from Answering |
| Human Collaboration | Involve human reviewers in the workflow | Human Collaboration |
| Sentiment & Feedback | Analyze tone, sentiment, and feedback signals | Sentiment Analyzer, Feedback Sentiment Analyzer |
Choosing the Right Categorizer
The platform offers several classification agents. Use this guide to pick the right one:
| Agent | Best for | Input type | Key feature | Status |
|---|---|---|---|---|
| Categorizer | Most new workflows | Any (via named inputs) | Flexible multi-input classification with rules | Current |
| Categorize Conversations and Route | Chat-based AI Employees | Chat conversation | Multi-intent detection, human collaboration (ask user to pick), example query matching | Current |
| Thread Categorizer | Ticketing workflows | Support ticket thread | Reads ticket title, description, and recent comments | Current |
| Text Categorizer | Simple text classification | Plain text | Streamlined, text-only | Deprecated (use Categorizer) |
| Document Categorizer | Document classification | Documents | Document-only classification | Deprecated (use Categorizer) |
Rules of thumb:
- If your input is a chat conversation and you need intent routing, use Categorize Conversations and Route. It supports multi-intent detection and can ask the end user to choose when ambiguous.
- If your input is a support ticket, use Thread Categorizer. It understands ticket structure (title, description, comments).
- For everything else -- documents, emails, extracted data, plain text, or any combination -- use the Categorizer.
Core Agents
Core agents are general-purpose building blocks that apply across industries and use cases. Every Ema workspace has access to all core agents.
| Agent | Category | Description |
|---|---|---|
| Abstain from Answering | Validation & QA | Stops a workflow branch from producing a response when the AI Employee should not answer |
| Agentic Search and Respond | Search & Retrieval | Iterative research agent that plans searches, discovers file metadata, and synthesizes answers with citations |
| App Navigator | External Integration | Browser automation agent that navigates web applications and performs UI-level actions |
| Categorize Conversations and Route | Classification & Routing | Classifies chat conversations into categories and routes to the appropriate workflow branch. Supports multi-intent detection and human collaboration. |
| Categorizer | Classification & Routing | Flexible categorization of any input type using named inputs, deterministic rules, and LLM classification. Recommended for most new workflows. |
| Chart Builder | Document Generation | Generates charts and visualizations from tabular markdown data |
| Combine Search Results | Data Processing | Merges and deduplicates search results from multiple sources with configurable weighting |
| Combine Text | Data Processing | Merges multiple text inputs using LLM summarization or simple append mode |
| Conversation Summarizer | Data Processing | Converts multi-turn chat conversations into concise search queries. Supports both simple concatenation and LLM-powered summarization. |
| Convert to Text | Data Processing | Deterministic conversion of any well-known type to plain text |
| Custom Agent | Response Generation | General-purpose LLM agent with configurable inputs, outputs, and instructions |
| Custom Code Agent | Data Processing | Executes custom JavaScript in a sandboxed environment for data transformation, conditional logic, and LLM calls |
| Deep Web Search | Search & Retrieval | Multi-strategy web search with AI-powered result synthesis |
| Document Categorizer | Classification & Routing | Deprecated. Categorizes documents into predefined categories. Use the Categorizer with document named inputs instead. |
| Document Synthesis | Search & Retrieval | Agentic document search with image and table support |
| Extract Entities | Data Processing | Schema-driven entity extraction from documents and text |
| Feedback Sentiment Analyzer | Sentiment & Feedback | Classifies feedback as positive, negative, or no feedback detected |
| Fixed Response | Response Generation | Template-based responses with variable substitution from workflow data |
| Generate Document | Document Generation | Creates formatted .docx documents from markdown with style guide support |
| Human Collaboration | Human Collaboration | Routes decisions to human reviewers with conversational, standard form, or custom form modes |
| Intelligent Actions | External Integration | Executes actions across 200+ connected applications via tool calling |
| JSON Extractor | Data Processing | Extracts values from nested JSON structures into flat key-value pairs |
| Knowledge Search | Search & Retrieval | Searches document knowledge bases using semantic retrieval with configurable data sources |
| Mark as Public Response | Validation & QA | Sets a boolean flag indicating whether a response should be publicly visible on tickets |
| Rerank Search Results | Search & Retrieval | LLM-based reranking of combined search results for improved relevance |
| Respond to a Query | Response Generation | General-purpose response generation from any combination of inputs |
| Respond using Action Calling Results | Response Generation | Formats and narrates the outputs of external tool executions |
| Respond using Search Results | Response Generation | Generates responses grounded in search results with source citations |
| Response Validator | Validation & QA | Validates generated responses against configurable quality and compliance criteria |
| Rule Validation | Validation & QA | Validates documents or data against a defined rule set with pass/fail output |
| Sales Intelligence | External Integration | Retrieves financial and company data from external APIs |
| Send Email | External Integration | Sends emails with attachment support, CC/BCC, HTML formatting, and read tracking |
| Sentiment Analyzer | Sentiment & Feedback | Detailed sentiment classification with nuance and sub-category detection |
| Tag Extractor | Data Processing | Extracts tags from conversations using configurable mapping rules |
| Text Categorizer | Classification & Routing | Deprecated. AI-powered text classification. Use the Categorizer for new workflows. |
| Thread Categorizer | Classification & Routing | Categorizes support ticket threads based on title, description, and recent comments |
| Thread Summarizer | Data Processing | Summarizes support ticket threads into concise overviews covering issue, actions taken, and status |
Specialized Agents
Specialized agents are pre-trained for specific industry or domain tasks. They encapsulate domain expertise and are typically used as drop-in replacements for the Custom Agent when a particular use case is well-defined.
See the Specialized Agents Catalog for the full list.
Deprecated Agents
Some agents have been superseded by newer implementations. See Deprecated Agents for the list and migration guidance.
For internal implementation details (action registration, knowledge layer, skill bundles, and agent QA pipeline), see the internal agent reference supplement.