1. Overview

The BOOST API is designed for controlled, server-side integrations—allowing you or your technical team to call BOOST tools from your own systems while maintaining the same Fair Housing and CREA-aware safeguards as the web interface.

The API surface will be implemented on Cloudflare Workers and exposed behind a stable HTTPS base URL. A Machine Config Protocol (MCP) endpoint will be available for AI clients that support MCP.

All examples and paths on this page are representative. Exact hostnames, paths, and fields will be finalized when the API is launched.

2. Authentication

API access will use secret keys issued from your BOOST account. Keys will be tied to a specific workspace or user and inherit that account's plan, usage limits, and allowed tools.

• API keys must be kept confidential and never embedded in client-side code.
• All requests must use HTTPS and include an Authorization header.
• Keys can be rotated or revoked from the BOOST dashboard (Billing & Subscription or API Access sections).

Authorization: Bearer YOUR_API_KEY

3. Tool endpoints

Each BOOST tool will be exposed as a structured endpoint that accepts JSON input and returns JSON output. Inputs and outputs are designed to stay within safe criteria and avoid protected-class content.

Representative pattern for tool execution:

POST /v1/tools/<toolId>/generate

Representative request body:

{
  "context": {
    "agentRole": "listing-agent",
    "locale": "en-US"
  },
  "input": {
    "notes": "Neutral property description and client goals...",
    "tone": "professional",
    "channel": "email"
  }
}

Representative response body:

{
  "id": "tool-run-123",
  "toolId": "listing-description",
  "output": {
    "text": "Generated, compliance-aware content...",
    "tokensUsed": 123
  },
  "meta": {
    "plan": "pro-bundle-example",
    "usageRecorded": true
  }
}

Response text will be generated with safeguards that avoid crime, school ratings, demographic targeting, and neighborhood “quality” labels. Not legal advice. Not financial advice.

4. MCP integration

For AI clients that support the Machine Config Protocol (MCP), BOOST will expose a server that lists tools and schemas in a machine-readable format. This allows AI agents to call BOOST tools under your account while respecting usage limits and compliance filters.

• MCP will describe each tool, accepted fields, and safe usage patterns.
• Only approved tools and methods will be available via MCP.
• MCP access will still require authentication, using a key or configuration that maps to your BOOST account.

5. Usage metering and limits

The API will use the same usage-based metering as the web application. Each request is counted against your plan's limits, whether it originates from the dashboard, a Worker, or an MCP integration.

• Usage may be tracked per tool, per month, and per account or workspace.
• Exceeded limits will result in clear, structured error responses.
• Usage details will be visible in your BOOST dashboard under Billing & Subscription and Dashboard > Usage overview.

{
  "error": {
    "code": "usage_limit_exceeded",
    "message": "Your current plan limit for this tool has been reached.",
    "plan": "pro-bundle-example",
    "toolId": "listing-description"
  }
}

6. Webhooks for billing and usage events

When billing and subscription features are enabled, BOOST will send webhook events for subscription changes, plan updates, and relevant usage thresholds. These events allow your systems to stay in sync with your BOOST account.

• Events are delivered to HTTPS endpoints you configure.
• Each event includes a type, timestamp, and signed payload.
• Signatures must be verified before processing the event.

{
  "id": "evt_123",
  "type": "subscription.updated",
  "created": "2025-01-01T12:00:00Z",
  "data": {
    "plan": "pro-bundle-example",
    "status": "active"
  }
}

Webhook events are informational. Final billing and tax obligations should always be confirmed with your billing provider and professional advisors. Not legal advice. Not financial advice.

7. Compliance considerations for API use

When you integrate with the BOOST API, you are responsible for ensuring that prompts and outputs used in your systems remain aligned with Fair Housing and CREA rules.

• Do not send prompts that request crime statistics, school ratings, or demographic targeting.
• Do not build automated flows that segment clients or leads by protected-class criteria or neighborhood “quality.”
• Always review generated content before sending it to clients or publishing it on public channels.

Fair Housing/CREA compliant; no protected-class criteria used. Not legal advice. Not financial advice. Always verify with your broker, brokerage policies, and applicable authorities.