Work with this as data
Every API here is available over the APIs.io API and to AI agents over MCP.
MCP server
One button, every client — Claude, Cursor, VS Code and the rest.
https://apis.io/mcp
Tools for apis
7 MCP tools reach this
find_apisBrowse and filter every API in the catalog.get_api_artifactsOne API's artifacts, grouped by type.get_openapiThe primary OpenAPI for this API.find_similar_apisAPIs that look like this one.apis_io_searchSTART HERE — APIs, providers and tags for one query, each with its total.resolveTurn a domain, URL or GitHub org into the provider it belongs to.find_cohortsEvery scored population of providers in the catalog.
Call it yourself
curl for this page
This API
curl "https://apis.io/api/v1/apis/mabl:mabl-chat-api"
All apis
curl "https://apis.io/api/v1/apis?limit=25"
Calls without X-API-Key need no key. Calls with it need a paid plan — get a key.
Get an API key
Free tier, no form to fill in. Signing in shares your email address with us — we store it to create your key and to recognise you if you sign in with another provider. See our Privacy Policy and Terms.
A second provider on the same verified email joins the account you already have.
OpenAPI Specification
openapi: 3.2.0
info:
title: mabl Chat API
description: APIs for interacting with mabl services
version: 0.599.3.1792436330
x-readme-id: 62c896636916ce01d2ff4f5a
servers:
- url: /
security:
- basicAuth: []
- bearerAuth: []
tags:
- name: Chat
paths:
/chat/{workspace_id}/request:
post:
tags:
- Chat
summary: Make a chat request
description: Make a chat request to the assistant
operationId: requestChat
parameters:
- name: workspace_id
in: path
description: The workspace associated with this assertion
required: true
style: simple
explode: false
schema:
type: string
requestBody:
description: A chat request to the assistant including relevant user context and chat history
content:
application/json:
schema:
$ref: '#/components/schemas/ChatRequest'
required: true
responses:
'200':
description: A response from the assistant to the user's request
content:
application/json:
schema:
$ref: '#/components/schemas/ChatResponse'
'400':
$ref: '#/components/responses/ErrorResponse400'
'401':
$ref: '#/components/responses/ErrorResponse401'
'403':
$ref: '#/components/responses/ErrorResponse403'
default:
$ref: '#/components/responses/ErrorResponse'
x-mabl-action: Run::Chat
components:
schemas:
EntityTypeEnum:
type: string
enum:
- application
- branch
- datatable
- deployment_event
- environment
- file_upload
- flow
- insight
- plan
- plan_run
- snippet
- test
- test_run
ChatRequest:
required:
- user_context
- user_text
type: object
properties:
user_text:
type: string
description: The user's latest text input to the assistant
session_id:
type: string
description: The ID of the current conversation session for tracking purposes
user_context:
$ref: '#/components/schemas/ChatUserContext'
history:
type: array
description: The chat history leading up to this request
items:
$ref: '#/components/schemas/ChatHistoryItem'
module_override:
type: string
description: Temporarily force router to use specific module for hackathon only. String must match getName() value of the module
ChatResponse:
required:
- response_text
- workspace_id
type: object
properties:
workspace_id:
type: string
description: The workspace associated with this assertion evaluation
response_text:
type: string
description: The assistant's response to the user's request
action_suggestions:
type: array
description: Suggestions for actions the user can take next
items:
$ref: '#/components/schemas/ChatAction'
entities:
type: array
description: Entities, by ID, that are referenced in the assistant's response and provide context in subsequent conversation or action suggestions
items:
$ref: '#/components/schemas/ChatContextEntityList'
GuardrailApprovalKind:
type: string
description: Who decides a guardrail approval. SELF is decided by the person the agent acts for; WORKSPACE, for an agent acting for an API key, by a workspace owner or editor.
enum:
- SELF
- WORKSPACE
ChatContextEntityList:
required:
- entity_type
- ids
- is_focus
type: object
properties:
entity_type:
$ref: '#/components/schemas/EntityTypeEnum'
is_focus:
type: boolean
description: Whether the entity is the primary focus (for that entity type) in the user's current context. If true, the list of IDs should contain only one ID.
ids:
type: array
description: List of entity IDs
items:
type: string
names:
type: array
description: Optional list of entity names by same index as IDs
items:
type: string
ChatFunctionCall:
required:
- function_name
type: object
properties:
function_name:
type: string
description: The name of the function that was called
arguments:
type: object
additionalProperties: true
description: The arguments passed to the function
output:
type: object
additionalProperties: true
description: The output returned by the function
succeeded:
type: boolean
description: Whether the function call succeeded (may be undefined while in progress during streaming)
error_message:
type: string
description: Error message if the function call failed
description: A function call made during a chat message. Function calls are only supported for model messages.
ChatMessageRoleEnum:
type: string
enum:
- user
- model
GuardrailApprovalRef:
type: object
properties:
id:
type: string
description: The approval's id.
state:
description: Where the approval stands.
allOf:
- $ref: '#/components/schemas/GuardrailApprovalState'
kind:
description: Who decides the approval.
allOf:
- $ref: '#/components/schemas/GuardrailApprovalKind'
request_expires_time:
type: integer
description: When the approval expires if nobody decides it, in epoch milliseconds.
format: int64
url:
type: string
description: The mabl app page where a person reviews the approval.
description: 'An approval a guardrail policy asks for. While it is PENDING, stop: give the person the url, tell them the request is waiting for approval, and do not repeat the request or poll in a loop. Once a person tells you it is APPROVED, repeat the identical request with the header X-Mabl-Approval set to id. DENIED means a person said no, so do not work around it.'
ChatAction:
required:
- action_type
- display_text
type: object
properties:
action_type:
type: string
description: The type of action to suggest to the user
enum:
- deep_link
- external_link
- failure_categorization
- create_plan_from_test_list
- add_label
display_text:
type: string
description: The text to display for the external link suggested action
discriminator:
propertyName: action_type
mapping:
deep_link: '#/components/schemas/ChatActionDeepLink'
external_link: '#/components/schemas/ChatActionExternalLink'
failure_categorization: '#/components/schemas/ChatActionFailureCategorization'
create_plan_from_test_list: '#/components/schemas/ChatActionCreatePlanFromTestList'
add_label: '#/components/schemas/ChatActionAddLabel'
ChatHistoryItem:
required:
- role
type: object
properties:
role:
$ref: '#/components/schemas/ChatMessageRoleEnum'
message:
type: string
description: The text of the chat message (may be partial or absent during streaming)
timestamp:
type: integer
description: The time the item was generated in epoch millis
format: int64
function_call_groups:
type: array
description: Function calls made during this chat message. Function calls are only supported for model messages. Each item is a group of function calls made in parallel, and following function calls were executed serially after the previous function calls.
items:
$ref: '#/components/schemas/ChatFunctionCallGroup'
streamed_thinking:
type: string
description: The model's thinking output that is being streamed in real-time (only populated during streaming, accumulates until next function call round)
thinking:
type: string
description: The model's final thinking output for this message
ChatFunctionCallGroup:
required:
- function_calls
type: object
properties:
function_calls:
type: array
description: Contains a list of function calls made in parallel.
items:
$ref: '#/components/schemas/ChatFunctionCall'
text:
type: string
description: The text output from the model prior to generating this group of function calls
thinking:
type: string
description: The model's thinking output that led to generating this group of function calls
ErrorResponseGeneric:
type: object
properties:
id:
type: string
x-mabl-scope: public
code:
maximum: 599
minimum: 100
type: integer
format: int32
x-mabl-scope: public
message:
type: string
x-mabl-scope: public
error_code:
type: string
description: Machine-readable reason for the error, present only when the API names one. Values are guardrail_blocked (a guardrail policy refused an agent request), approval_required (a guardrail policy requires a person to approve the request first).
guardrail:
description: The guardrail policy that refused the request; present only when error_code is guardrail_blocked or approval_required.
allOf:
- $ref: '#/components/schemas/GuardrailRefusal'
GuardrailRefusal:
type: object
properties:
id:
type: string
description: The id of the guardrail policy that decided the refusal.
reason:
type: string
description: Why the policy refuses the request, written for the person the agent acts for.
decision_id:
type: string
description: The id of this decision in the guardrail decision log.
action:
type: string
description: The action the request was evaluated as, written Verb::Resource (for example Delete::Test), as declared by the operation's x-mabl-action extension.
policy_source:
type: string
description: 'Who set the deciding policy: workspace for a rule the workspace''s owners wrote, account for a rule set for the workspace''s account, mabl_agent_type for the agent type''s ceiling, and mabl_default for mabl''s own defaults.'
approval:
description: The approval for this request; present when error_code is approval_required and the approval could be recorded. Without it the request cannot be approved.
allOf:
- $ref: '#/components/schemas/GuardrailApprovalRef'
description: The guardrail policy that refused an agent request.
GuardrailApprovalState:
type: string
description: Where a guardrail approval stands. PENDING awaits a decision and expires undecided after 24 hours; APPROVED lets the agent repeat the request, as often as it does, for 20 minutes; DENIED refuses the same request for 20 minutes after the decision, after which asking again opens a new approval; CANCELLED means a person withdrew it while pending or revoked it while approved. EXPIRED and CANCELLED are final.
enum:
- PENDING
- APPROVED
- DENIED
- EXPIRED
- CANCELLED
ChatUserContext:
required:
- location_description
type: object
properties:
location_description:
type: string
description: Short description of the location/source from which the chat request was made, such as the current page or view in the app or from an integration
additional_context_description:
type: string
description: Additional, detailed, free-text context about the user's current context in the application, such as selected/highlighted information or visible data not easily retrievable on the backend
entities:
type: array
description: Entities, by ID, relevant to current user context (especially any entities currently shown in UI)
items:
$ref: '#/components/schemas/ChatContextEntityList'
responses:
ErrorResponse400:
description: Invalid or missing parameter
content:
application/json:
schema:
$ref: '#/components/schemas/ErrorResponseGeneric'
ErrorResponse403:
description: User not authorized
content:
application/json:
schema:
$ref: '#/components/schemas/ErrorResponseGeneric'
ErrorResponse:
description: Unknown error
content:
application/json:
schema:
$ref: '#/components/schemas/ErrorResponseGeneric'
ErrorResponse401:
description: User not authenticated
content:
application/json:
schema:
$ref: '#/components/schemas/ErrorResponseGeneric'
securitySchemes:
basicAuth:
type: http
scheme: basic
bearerAuth:
type: apiKey
name: Authorization
in: header