create tool

Create a new tool for function calling in persona sessions

POST /v1/tools

cURL

default

curl --request POST \
  --url https://api.anam.ai/v1/tools \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "name": "open_calendar",
  "description": "Open the calendar UI in the client app.",
  "type": "CLIENT",
  "config": {
    "parameters": {
      "type": "object",
      "properties": {
        "date": {
          "type": "string",
          "description": "Date to jump to, in YYYY-MM-DD format."
        }
      }
    }
  }
}
'

201

{
  "id": "00000000-0000-0000-0000-000000000000",
  "name": "open_calendar",
  "description": "Open the calendar UI in the client app.",
  "type": "CLIENT",
  "config": {
    "parameters": {
      "type": "object",
      "properties": {
        "date": {
          "type": "string",
          "description": "Date to jump to, in YYYY-MM-DD format."
        }
      }
    }
  },
  "createdAt": "2026-04-20T10:00:00.000Z",
  "updatedAt": null,
  "usageCount": 0
}

Authorizations

Authorization · string · header · required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Body

application/json

Tool definition. config depends on type. For example: {"name":"lookup_order","description":"Look up an order by its number","type":"SERVER_WEBHOOK","config":{"url":"https://api.example.com/orders","method":"POST","awaitResponse":true,"parameters":{"type":"object","properties":{"orderId":{"type":"string","description":"The caller's order number"}},"required":["orderId"]}}}.

name · string · required

Name the LLM calls the tool by. Characters other than letters, digits, _, . and - (spaces included) are saved as _, with repeated _ merged and any at either end dropped; names starting with anam_ are reserved. Names needn't be unique: when a persona has two tools with one name, its sessions add _2, _3 and so on.

Required string length: 1 - 64

Example:

"search_knowledge_base"

description · string · required

Description of what the tool does. Used by the LLM to decide when to call it.

Required string length: 1 - 1024

Example:

"Search the knowledge base for information about products and services"

type · enum<string> · required

CLIENT sends the call to the client SDK, SERVER_RAG searches knowledge folders and SERVER_WEBHOOK calls a URL. SYSTEM tools (end_call, interrupt) are Anam's own and can't be created.

Available options: CLIENT, SERVER_RAG, SERVER_WEBHOOK

Example:

"SERVER_RAG"

disableInterruptions · boolean · default: false

The user can't interrupt the persona while this tool runs. Defaults to false.

Example:

false

config · CLIENT · object

Settings in the shape for the tool's type; fields of other shapes are stored but not used. null saves an empty config.

CLIENT

SERVER_RAG

SERVER_WEBHOOK

Show child attributes

Response

201

application/json

Successfully created tool

id · string<uuid>

Unique identifier for the tool

Example:

"00000000-0000-0000-0000-000000000000"

name · string

Name of the tool

Example:

"search_knowledge_base"

description · string

Description of what the tool does

Example:

"Search the knowledge base for product information"

type · enum<string>

Type of tool

Available options: CLIENT, SERVER_RAG, SERVER_WEBHOOK, SYSTEM

config · object

Type-specific configuration

disableInterruptions · boolean · default: false

When true, interruptions are disabled while this tool is executing

createdAt · string<date-time>

When the tool was created

updatedAt · string<date-time> | null

When the tool was last updated

usageCount · integer

Number of the organization's live personas using this tool