Skip to main content

Claude Code

Use tool_use to call the endpoints below directly from Claude Code.

Cursor

Cursor’s agent mode can call these endpoints via HTTP requests.

OpenClaw

Any tool-use-capable LLM agent can complete this flow.
These endpoints let AI agents set up Topify.ai from scratch and generate a full GEO visibility report using account details explicitly supplied or authorized by the user. The full flow is three API calls:
  1. Create account — register with an email, receive credentials instantly
  2. Get API key — exchange credentials for an API key
  3. Create project — start brand tracking and receive results via webhook
Use an email address supplied or explicitly authorized by the user. Agents must not invent a temporary email or admin identity. API keys can only be created by admins and currently grant access to the full team; project-scoped API keys are not available.
Run this flow only in a trusted server-side environment. Never expose the generated password or API key in browser code, client-side applications, logs, prompts, chat history, or tool output shown to other users.
These endpoints are designed for tool-use AI agents such as Claude Code, Cursor, and OpenClaw that can make HTTP requests programmatically. If you are using a web-based AI chatbot (ChatGPT, Perplexity, etc.) that cannot call APIs directly, create your account and project manually at app.topify.ai, then use the Get API key endpoint with your email and password to obtain a key for the read-only data endpoints.

Credits and rate limits

Every API key has a rate limit tier and a monthly credit allowance. Credits are consumed when projects fetch AI responses during bootstrap and daily refresh cycles.

Rate limit tiers

Rate limits use a sliding 60-second window. The burst allowance lets you briefly exceed the base rate. New API keys are created on the standard tier. Contact us to upgrade to premium. When the limit is exceeded, the API returns 429 Too Many Requests with a Retry-After header indicating how many seconds to wait.

Research credits by plan

Each plan includes a monthly credit budget that determines how many AI research queries your projects can run. Accounts created via the API start on the skip_trial plan with 10 credits. Each project bootstrap consumes credits proportional to the number of prompts and AI providers queried (typically 5 prompts across 3 providers = 15 credits per full bootstrap cycle). The skip_trial plan runs a reduced bootstrap that fits within the 10-credit budget. For a full bootstrap with all prompts and providers, upgrade to the Basic plan or higher from the dashboard or contact us.

Response headers

Every response includes rate limit and credit information:

Create account

Creates a new Topify.ai account with a team on the skip_trial plan. The account is created without email verification, so it can be used immediately.
No authentication is required for this endpoint. The generated password is returned once and cannot be retrieved later — store it securely.

Request body

Response

Response fields

Errors

If you receive a 409: The account already exists. Skip to the Get API key step using the same email and the user’s existing password. If the password is unknown, ask the user or direct them to reset it at app.topify.ai.

Get API key

Authenticates a team administrator with email and password, then creates and returns a new API key on the standard rate limit tier. The key can access every project owned by the selected team; project-scoped keys are not currently issued.

Request body

Response

Response fields

Errors


Create project

Creates a new brand tracking project and starts the bootstrap pipeline in the background. The endpoint returns immediately with a 202 Accepted status while the pipeline runs asynchronously. The bootstrap pipeline generates tracking prompts, fetches initial AI responses from all providers, calculates brand metrics, and detects competitors. When complete, a webhook callback is sent to the URL you provide.
Requires API key authentication via the X-API-Key header.

Request body

Response

Response fields

Webhook callback

When the bootstrap pipeline completes (or fails), a POST request is sent to your webhook_url with the following payload: Success:
Failure:
Treat the bootstrap callback only as a completion signal. Match its project_id to the ID returned by your create request, then confirm the current project state through an authenticated API request before performing any follow-up mutation.

Errors


After bootstrap completes

After the webhook reports "status": "completed" and you confirm the project through the authenticated API, use these read-only endpoints (documented in API reference) to retrieve results: All data endpoints require the X-API-Key header and return responses in the format {"success": true, "data": {...}}.
For a quick summary to present to the user, start with GET /projects to get the brand’s overall visibility score and sentiment, then GET /projects/{id}/overview for a prompt-level breakdown.

Full workflow example

The workflow can run end-to-end after the user supplies or explicitly authorizes the account email. Do not generate a temporary identity. The API key created in step 2 is an admin-created, full-team credential.