Agent

Wabee Agent Core API (v1)

Wabee Agent Core API

Wabee Agent Core API enables developers to interact with an AI agent in a custom, secure and flexible manner through REST API calls.

One can build applications to interact with the agent API to complete a task using any of the Agents endpoints which allow for both text streaming and direct JSON response. The Memory endpoints are useful for managing the agent memory programatically. Moreover, the Metrics endpoints provide an interface for monitoring the underlying Agent in terms of latency, health, token consumption and much more.

Download OpenAPI description
openapi.json
openapi.yaml
Overview
License Wabee License
Languages
Servers
Mock server
https://api.docs.wabee.ai/_mock/openapi/
Production server
https://<your_agent_uri>.wabee.ai/

Agent

Endpoints for interacting with the AI agent

Operations

Demand from the agent and get a full response when it finishes

Request

Executes a chain of AI-powered reasoning and actions to respond to a given query or task.

The agent leverages its configured tools and knowledge to:

  • Analyze the input query
  • Plan necessary steps
  • Execute relevant tools
  • Synthesize a comprehensive response

Image Support

Include images in your request using the images field:

  • Base64 Format: Set type="base64" with base64-encoded image data
  • File Path: Set type="file_path" with path to existing image

Maximum 5 images per request. Invalid images will trigger a 400 response.

Authentication

Requires API key via x-wabee-access header.

Session Management

Use session-id header to maintain conversation context across requests.

Headers
session-idany(Session-Id)

Session unique identifier, used to keep track of previous interations and much more

Bodyapplication/jsonrequired
messagesArray of objects(Messages)required

List of messages to be processed by the agent

Example: [{"content":"What is the economic outlook?","role":"user"}]
messages[].​rolestring(Role)required

defines the role associated with the message

Example: "assistant"
messages[].​contentstring(Content)required

message content

Example: "Hello, I am your virtual assistant"
data_filter_hintsArray of Data Filter Hints (objects) or Data Filter Hints (null)(Data Filter Hints)

Special metadata-like hints to be passed to the agent in order to instruct him on how to filter the data on some tools that support filter

Example: [{"filter_key":"unitname","filter_type":"equals","filter_value":"Unit 1","tool_name":"bubble_data"}]
Any of:

Special metadata-like hints to be passed to the agent in order to instruct him on how to filter the data on some tools that support filter

context_filesArray of Context Files (strings) or Context Files (null)(Context Files)

The path for files which the agent already has access to and should be considered by it when processing the messages

Example: ["inputs/chat-files/test123/file.pdf"]
Any of:

The path for files which the agent already has access to and should be considered by it when processing the messages

imagesArray of Images (objects) or Images (null)(Images)

List of images associated with the message

Any of:

List of images associated with the message

curl -i -X POST \
  https://api.docs.wabee.ai/_mock/openapi/core/v1/chain \
  -H 'Content-Type: application/json' \
  -H 'session-id: ' \
  -H 'x-wabee-access: YOUR_API_KEY_HERE' \
  -d '{
    "messages": [
      {
        "content": "What is the economic outlook?",
        "role": "user"
      }
    ],
    "data_filter_hints": [
      {
        "filter_key": "unitname",
        "filter_type": "equals",
        "filter_value": "Unit 1",
        "tool_name": "bubble_data"
      }
    ],
    "context_files": [
      "inputs/chat-files/test123/file.pdf"
    ],
    "images": [
      {
        "content": "iVBORw0KGgoAAAANSUhEUgAAABgAAAAYCAYAAADgdz34AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOydd3gU5f7/3+9z",
        "type": "base64"
      }
    ]
  }'

Responses

Successful Response

Headers
session-idany

session unique identifier

request-idany

an auto-generated request unique identifier

Bodyapplication/json
bodyobject(ChainResponseModelBody)required

Represents the complete response from the AI agent

body.​inputstring(Input)required

The original input message sent to the agent

Example: "What is the economic outlook?"
body.​outputstring(Output)required

The final response generated by the agent

Example: "The economic outlook varies significantly by region..."
body.​intermediate_stepsArray of ChainResponseModelBodyIntermediateStep (object) or strings(Intermediate Steps)required

The sequence of steps taken by the agent to generate the response, including tool usage and reasoning

Example: [[{"log":"Retrieving news data","tool":"api_tool","tool_input":"curl https://newsapi.org/v2/everything","type":"AgentAction"},"Explanation: Gathering real-time economic data from trusted sources"]]
Array [
Any of:

Represents a single step in the agent's reasoning process

body.​intermediate_steps[][].​toolstring(Tool)required

The identifier of the tool used in this step

Example: "api_tool"
body.​intermediate_steps[][].​tool_inputstring(Tool Input)required

The input provided to the tool

Example: "curl https://newsapi.org/v2/everything"
body.​intermediate_steps[][].​logstring(Log)required

Detailed log of the tool execution

Example: "Retrieving news data"
body.​intermediate_steps[][].​typestring(Type)required

The type of action performed in this step

Example: "AgentAction"
]
body.​usageArray of objects(Usage)required

Resource usage metrics for the request

body.​usage[].​costAttributestring(Costattribute)required

The name of the resource being measured (e.g. model name)

Example: "gpt-4-32k"
body.​usage[].​usagestring(Usage)required

The quantity of the resource used

Example: "100"
body.​usage[].​unitstring(Unit)required

The unit of measurement (e.g. tokens, requests)

Example: "tokens"
Response
application/json
{
  "body": {
    "input": "What is the economic outlook?",
    "output": "The economic outlook varies significantly by region...",
    "intermediate_steps": [ … ],
    "usage": [ … ]
  }
}

Demand from the agent and get a streamed response until it finishes

Request

Requests Wabee AI Agent to answer a question regarding the mission that the agent was given to upon creation.

The agent can make use of any of the available tools along with its own reasoning to complete the given task, depending on its complexity, the response might take some time to be returned to the user.

Note: session_id field on request body is kept for compatibility purposes, it will be overwritten with the header value during request processing Note: stream filed on request body is kept for compatibility purposes, it will be ignored during request processing Note: this is pretty much the same as in /chain, the only difference is that this endpoint returns a streaming response instead of a JSON object Note: the id on the streamming response refers to the current request Note: the streamming response ends when finish_reason is different than null Note: All headers starting with stix will be sent as context for tools.

Headers
session-idany(Session-Id)

Session unique identifier, used to keep track of previous interations and much more

Bodyapplication/jsonrequired
messagesArray of objects(Messages)required

List of messages to be processed by the agent

Example: [{"content":"What is the economic outlook?","role":"user"}]
messages[].​rolestring(Role)required

defines the role associated with the message

Example: "assistant"
messages[].​contentstring(Content)required

message content

Example: "Hello, I am your virtual assistant"
data_filter_hintsArray of Data Filter Hints (objects) or Data Filter Hints (null)(Data Filter Hints)

Special metadata-like hints to be passed to the agent in order to instruct him on how to filter the data on some tools that support filter

Example: [{"filter_key":"unitname","filter_type":"equals","filter_value":"Unit 1","tool_name":"bubble_data"}]
Any of:

Special metadata-like hints to be passed to the agent in order to instruct him on how to filter the data on some tools that support filter

context_filesArray of Context Files (strings) or Context Files (null)(Context Files)

The path for files which the agent already has access to and should be considered by it when processing the messages

Example: ["inputs/chat-files/test123/file.pdf"]
Any of:

The path for files which the agent already has access to and should be considered by it when processing the messages

imagesArray of Images (objects) or Images (null)(Images)

List of images associated with the message

Any of:

List of images associated with the message

curl -i -X POST \
  https://api.docs.wabee.ai/_mock/openapi/core/v1/chain_streaming \
  -H 'Content-Type: application/json' \
  -H 'session-id: ' \
  -H 'x-wabee-access: YOUR_API_KEY_HERE' \
  -d '{
    "messages": [
      {
        "content": "What is the economic outlook?",
        "role": "user"
      }
    ],
    "data_filter_hints": [
      {
        "filter_key": "unitname",
        "filter_type": "equals",
        "filter_value": "Unit 1",
        "tool_name": "bubble_data"
      }
    ],
    "context_files": [
      "inputs/chat-files/test123/file.pdf"
    ],
    "images": [
      {
        "content": "iVBORw0KGgoAAAANSUhEUgAAABgAAAAYCAYAAADgdz34AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOydd3gU5f7/3+9z",
        "type": "base64"
      }
    ]
  }'

Responses

Successful Response

Headers
session-idany

session unique identifier

request-idany

an auto-generated request unique identifier

Bodytext/event-stream
Response
text/event-stream
data: {"id": "3f364df0-a4e5-41b1-bb33-eeaa616a8060", "token": "A perspectiva econômica pode ", "agent_step": "PERGUNTA", "finish_reason": "null", "usage": [{"costAttribute": "gpt-4-32k", "usage": 100, "unit": "tokens"}]}

Demand from the agent and get a fast ack response, use callback to receive the completion response

Request

Requests Wabee AI Agent to answer a question asynchronously. If a callback_url is provided, the agent will send the response to the given URL.

The agent can make use of any of the available tools along with its own reasoning to complete the given task, depending on its complexity, the response might take some time to be returned to the user.

Note: session_id field on request body is kept for compatibility purposes, it will be overwritten with the header value during request processing Note: stream filed on request body is kept for compatibility purposes, it will be ignored during request processing

Bodyapplication/jsonrequired
specversionstring(Specversion)required

The version of the CloudEvents spec.

Example: "1.0"
typestring(Type)

The type of the event. Use the value 'ai.wabee.agent.chain' for chain requests.

Default "ai.wabee.agent.chain"
sourcestring(Source)required

The source of the event.

Example: "https://wabee.ai/agent"
idstring(Id)required

A unique identifier for the event.

Example: "A234-1234-1234"
timestring(Time)required

The time the event was created.

Example: "2018-04-05T17:31:00Z"
dataobject(MessageInputModelWithCallback)required

The data of the event.

data.​messagesArray of objects(Messages)required

List of messages to be processed by the agent

Example: [{"content":"What is the economic outlook?","role":"user"}]
data.​messages[].​rolestring(Role)required

defines the role associated with the message

Example: "assistant"
data.​messages[].​contentstring(Content)required

message content

Example: "Hello, I am your virtual assistant"
data.​data_filter_hintsArray of Data Filter Hints (objects) or Data Filter Hints (null)(Data Filter Hints)

Special metadata-like hints to be passed to the agent in order to instruct him on how to filter the data on some tools that support filter

Example: [{"filter_key":"unitname","filter_type":"equals","filter_value":"Unit 1","tool_name":"bubble_data"}]
Any of:

Special metadata-like hints to be passed to the agent in order to instruct him on how to filter the data on some tools that support filter

data.​context_filesArray of Context Files (strings) or Context Files (null)(Context Files)

The path for files which the agent already has access to and should be considered by it when processing the messages

Example: ["inputs/chat-files/test123/file.pdf"]
Any of:

The path for files which the agent already has access to and should be considered by it when processing the messages

data.​imagesArray of Images (objects) or Images (null)(Images)

List of images associated with the message

Any of:

List of images associated with the message

data.​session_idSession Id (string) or Session Id (null)(Session Id)

An optional session ID to associate with the chain. This can be used to track multiple interactions with the chain across different requests. If not provided, a session ID will be generated.

Any of:

An optional session ID to associate with the chain. This can be used to track multiple interactions with the chain across different requests. If not provided, a session ID will be generated.

string(Session Id)

An optional session ID to associate with the chain. This can be used to track multiple interactions with the chain across different requests. If not provided, a session ID will be generated.

data.​callback_urlCallback Url (string) or Callback Url (null)(Callback Url)

An optional callbacik URL to send the response to.

Any of:

An optional callbacik URL to send the response to.

string(Callback Url)

An optional callbacik URL to send the response to.

curl -i -X POST \
  https://api.docs.wabee.ai/_mock/openapi/core/v1/chain_async \
  -H 'Content-Type: application/json' \
  -H 'x-wabee-access: YOUR_API_KEY_HERE' \
  -d '{
    "specversion": "1.0",
    "type": "ai.wabee.agent.chain",
    "source": "https://wabee.ai/agent",
    "id": "A234-1234-1234",
    "time": "2018-04-05T17:31:00Z",
    "data": {
      "messages": [
        {
          "content": "What is the economic outlook?",
          "role": "user"
        }
      ],
      "data_filter_hints": [
        {
          "filter_key": "unitname",
          "filter_type": "equals",
          "filter_value": "Unit 1",
          "tool_name": "bubble_data"
        }
      ],
      "context_files": [
        "inputs/chat-files/test123/file.pdf"
      ],
      "images": [
        {
          "content": "iVBORw0KGgoAAAANSUhEUgAAABgAAAAYCAYAAADgdz34AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOydd3gU5f7/3+9z",
          "type": "base64"
        }
      ],
      "session_id": "string",
      "callback_url": "string"
    }
  }'

Responses

Successful Response

Bodyapplication/json
any
Response
application/json
null

Get wabee AI agent metadata

Request

Returns metadata information about the agent configuration, including its description, capabilities and other relevant information.

curl -i -X GET \
  https://api.docs.wabee.ai/_mock/openapi/core/v1/metadata \
  -H 'x-wabee-access: YOUR_API_KEY_HERE'

Responses

Successful Response

Bodyapplication/json
any
Response
application/json
{
  "name": "Wabee Agent",
  "description": "An AI agent that helps with specific tasks",
  "metadata": {}
}

Demand from the agent and get a streamed response until it finishesDeprecated

Request

Requests Wabee AI Agent to answer a question regarding the mission that the agent was given to upon creation.

The agent can make use of any of the available tools along with its own reasoning to complete the given task, depending on its complexity, the response might take some time to be returned to the user.

Images can be included in messages using the images field, which accepts a list of image inputs. Each image can be provided in one of two formats:

  • Base64-encoded strings: Set type="base64" and provide the base64-encoded image data in the content field
  • File paths: Set type="file_path" and provide the full path to the image file in the content field

Invalid image inputs will result in a 400 Bad Request response. The endpoint validates:

  • Base64 encoding format for base64 images
  • File existence for file path images

Note: session_id field on request body is kept for compatibility purposes, it will be overwritten with the header value during request processing Note: stream filed on request body is kept for compatibility purposes, it will be ignored during request processing Note: this is pretty much the same as in /chain, the only difference is that this endpoint returns a streaming response instead of a JSON object Note: the id on the streamming response refers to the current request Note: the streamming response ends when finish_reason is different than null Note: All headers starting with stix will be sent as context for tools.

Headers
session-idany(Session-Id)

Session unique identifier, used to keep track of previous interations and much more

Bodyapplication/jsonrequired
messagesArray of objects(Messages)required

List of messages to be processed by the agent

Example: [{"content":"What is the economic outlook?","role":"user"}]
messages[].​rolestring(Role)required

defines the role associated with the message

Example: "assistant"
messages[].​contentstring(Content)required

message content

Example: "Hello, I am your virtual assistant"
data_filter_hintsArray of Data Filter Hints (objects) or Data Filter Hints (null)(Data Filter Hints)

Special metadata-like hints to be passed to the agent in order to instruct him on how to filter the data on some tools that support filter

Example: [{"filter_key":"unitname","filter_type":"equals","filter_value":"Unit 1","tool_name":"bubble_data"}]
Any of:

Special metadata-like hints to be passed to the agent in order to instruct him on how to filter the data on some tools that support filter

context_filesArray of Context Files (strings) or Context Files (null)(Context Files)

The path for files which the agent already has access to and should be considered by it when processing the messages

Example: ["inputs/chat-files/test123/file.pdf"]
Any of:

The path for files which the agent already has access to and should be considered by it when processing the messages

imagesArray of Images (objects) or Images (null)(Images)

List of images associated with the message

Any of:

List of images associated with the message

curl -i -X POST \
  https://api.docs.wabee.ai/_mock/openapi/core/v1/chain_streamming \
  -H 'Content-Type: application/json' \
  -H 'session-id: ' \
  -H 'x-wabee-access: YOUR_API_KEY_HERE' \
  -d '{
    "messages": [
      {
        "content": "What is the economic outlook?",
        "role": "user"
      }
    ],
    "data_filter_hints": [
      {
        "filter_key": "unitname",
        "filter_type": "equals",
        "filter_value": "Unit 1",
        "tool_name": "bubble_data"
      }
    ],
    "context_files": [
      "inputs/chat-files/test123/file.pdf"
    ],
    "images": [
      {
        "content": "iVBORw0KGgoAAAANSUhEUgAAABgAAAAYCAYAAADgdz34AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOydd3gU5f7/3+9z",
        "type": "base64"
      }
    ]
  }'

Responses

Successful Response

Headers
session-idany

session unique identifier

request-idany

an auto-generated request unique identifier

Bodytext/event-stream
Response
text/event-stream
data: {"id": "3f364df0-a4e5-41b1-bb33-eeaa616a8060", "token": "The economic outlook may ", "agent_step": "QUESTION", "finish_reason": "null", "usage": [{"costAttribute": "gpt-4-32k", "usage": 100, "unit": "tokens"}]}

Memory

Endpoints for managing agent memory

Operations

Observability

Endpoints for monitoring agent performance and logs

Operations