ReferralCodes Agent Import 🦞

ReferralCodes supports AI agents and automation tools that can bulk upload referral codes directly to a user profile.

Agents can collect referral links and codes from sources such as email inboxes, purchase confirmations, referral dashboards and websites.

The official ReferralCodes referral finder skill can also be found here on ClawHub .


Agent Import Interface

Authenticated users can upload referrals using the agent import interface:

The interface supports two formats:

Agents should validate data before committing imports.

JSON Format

Agents should submit data using the following schema:

{
  "version": "1.0",
  "items": [
    {
      "shop": "brand name or domain",
      "discount": "reward for the new user",
      "url": "referral link (optional)",
      "code": "referral code (optional)",
      "description": "optional notes"
    }
  ]
}

Each item must include:


Example

{
  "version": "1.0",
  "items": [
    {
      "shop": "coinbase",
      "discount": "$10 in bitcoin",
      "url": "https://coinbase.com/join/example",
      "code": null,
      "description": "Coinbase referral bonus"
    },
    {
      "shop": "adidas",
      "discount": "15% off",
      "url": null,
      "code": "ADI15",
      "description": null
    }
  ]
}

Suggested Agent Workflow

  1. Search email inbox for referral messages.
  2. Extract referral links or codes.
  3. Convert results to the ReferralCodes JSON format.
  4. Open /profile/import/agent.
  5. Paste JSON and click Validate.
  6. If validation succeeds, click Commit.

Referral Extraction Guidance

Agents scanning inboxes should look for phrases such as:

Typical referral email subjects include:


Machine Readable Results

The agent import page exposes validation results in a machine-readable format for browser automation agents:

<script id="agent-import-result" type="application/json">

Agents can read validation results directly from the DOM after clicking Validate.


Future Interfaces


For questions about agent integrations contact the ReferralCodes team.

Referral extraction prompt