Skip to main content
Version: 3.x

Integrate the SDK with AI

Prefer to let an AI coding agent do the heavy lifting? Pick your platform and copy the prompt below into Claude, ChatGPT, Cursor, Codex — whatever you use. It points the agent at our full, always-current SDK reference for that platform, so it builds from the real documentation instead of guessing.

Copy this prompt into your AI assistant
I'm integrating TapResearch into my app using the SDK (iOS). Read the full reference documentation at https://supply-docs.tapresearch.com/llms-sdk-ios.txt first — it's the current, authoritative source — then help me implement it step by step, matching my project's existing structure and conventions. Follow the best-practices and testing guidance in the docs, and call out common mistakes as we go. If something isn't covered in the docs, tell me instead of guessing. Use placeholders like API_TOKEN for any credentials — don't ask me to paste real tokens or keys into this chat, and keep server-side keys out of client code and version control. Do not modify or create Info.plist for these — put the API token and placement tag in a separate Swift config file (e.g. TapResearchCredentials.swift). Show me where to add my API token and placement tag from the TapResearch dashboard. When the integration is in place, help me write and run a smoke test for each step and debug anything that fails — I'll confirm each step is passing in the TapResearch dashboard.

Your agent reads the live docs bundle, so it always has the current integration steps — no stale copy, nothing to keep in sync.

Open in ClaudeOpen in ChatGPT

Using a terminal or IDE agent (Claude Code, Cursor, Codex)? The Add to AGENTS.md line drops into your repo so the agent picks it up automatically. No web access in your agent? Open the full docs bundle and paste it in directly.

Your agent can scaffold initialization, reward callbacks, and placement display from there. You stay in control — review what it writes, and dive into the rest of these docs whenever you want the details yourself.

After you integrate: verify it's working​

Once the code is in place, run through these checks on a test device. Your AI agent can help — it can write a smoke test, walk you through each step, and debug failures. What it can't do is see your account, so your app's Integration tab in the Publisher Dashboard is what confirms each step actually passed — it detects each one from real activity and marks it done.

Use a real test session

These checks only register from a real test user reaching the SDK — firing a test postback on its own won't complete them. Add a test device first (More → Test devices); only test users can initialize the SDK before go-live.

  1. SDK initialization received — Initialize the SDK with your app's API token. The Integration tab marks this once it sees your first session.
  2. Enter survey wall from app — Show a placement and reach the survey wall. (You'll need a placement with content configured on the Manage tab first.)
  3. Reward callback configured — Set a server-to-server postback URL (this auto-completes), or implement the in-app reward handler.
  4. Generate test survey complete — Complete a test survey so a completion is recorded against your app.
  5. Reward redeemed — Confirm your callback or handler redeems the reward for the test user.

If a step doesn't check off, it's the fastest signal of where the integration broke. For deeper testing guidance, see Testing & QA.