pretestads-agent-scripts
Score video/image ads with the PreTestAds neural model from OpenAI, Claude, Gemini, LangChain, CrewAI, Vercel AI SDK, or any x402-capable agent — pay-per-request in USDC, no API key needed.
Documentation
PreTestAds Agent Scripts
Open-source scripts for scoring video/image ads with the PreTestAds neural model (AdCortex™, trained on fMRI brain-response data) — from any popular LLM platform using your own API key, or from any x402-capable agent with a crypto wallet.
No PreTestAds account. No PreTestAds API key. Each scoring run costs $5.00 USDC, paid per-request over the x402 protocol (HTTP 402) on Solana or Base.
What the API does
`POST https://pretestads.com/api/v1/score` — submit a video (MP4/MOV/AVI) or image (JPG/PNG/WebP) ad by URL or file upload. You get back a job id; 5–8 minutes later, `GET https://pretestads.com/api/v1/score/{id}` (free) returns:
- Score 0–100 (percentile vs. 76 top-performing TikTok ads) with a label: weak ( python score_ad.py https://example.com/ad.mp4
or a local file:
EVM_PRIVATE_KEY=0x... python score_ad.py ~/Desktop/ad.mp4
Node:cd javascript && npm install
EVM_PRIVATE_KEY=0x... node score-ad.mjs https://example.com/ad.mp4
## Quickstart (LLM agent with your API key)cd python && pip install -r requirements.txt openai
OPENAI_API_KEY=sk-... SOLANA_PRIVATE_KEY= \
python openai_agent.py "Score this ad and tell me how to improve it: https://example.com/ad.mp4"
The LLM decides when to call the `score_ad` tool; the tool pays $5 USDC via x402, waits for the result, and hands the score back to the model to interpret.
## How the x402 flow works
1. `POST /api/v1/score` with no payment → **HTTP 402** with an `accepts` array (price, USDC asset, network, payTo for Solana and Base).
2. Your x402 client SDK signs a USDC payment authorization with your local key.
3. Retry the POST with the payment header → server verifies + settles via the facilitator → **HTTP 202** with `{id, status_url, payment}`.
4. Poll `GET /api/v1/score/{id}` (free) until `status` is `complete` (~5–8 min).
One payment = one run. Replayed payment proofs are rejected. Invalid media (422) is never charged.
## Wallet setup
- **Solana**: a wallet holding USDC and a little SOL. Export the base58 secret key.
- **Base**: a wallet holding USDC and a little ETH on Base. Export the 0x private key.
- Use a dedicated hot wallet with only what you intend to spend — these scripts read the key from an environment variable.
## Environment variables
See [`.env.example`](.env.example). Never commit a real `.env`.
## License
MIT — see [LICENSE](LICENSE).Frequently asked questions
What is pretestads-agent-scripts?
pretestads-agent-scripts is Score video/image ads with the PreTestAds neural model from OpenAI, Claude, Gemini, LangChain, CrewAI, Vercel AI SDK, or any x402-capable agent — pay-per-request in USDC, no API key needed.
How do I install pretestads-agent-scripts?
Open the GitHub repository and follow its README. Most MCP servers are added to your client's MCP config, then called by your agent.
Is pretestads-agent-scripts open source?
Yes — it is hosted on GitHub at https://github.com/krecicki/pretestads-agent-scripts.
Related MCP tools
cogneeCognee is the open-source AI memory platform for agents. Give your AI agents persistent long-term memory across sessions with a self-hosted knowledge graph engine.
skyvernAutomate browser based workflows with AI
hindsightHindsight: Agent Memory That Learns
watermarks-removerA privacy-first app that strips AI watermarks from content you own.
Qwen-AgentAgent framework and applications built upon Qwen>=3.0, featuring Function Calling, MCP, Code Interpreter, RAG, Chrome extension, etc.
pal-mcp-serverThe power of Claude Code / GeminiCLI / CodexCLI + [Gemini / OpenAI / OpenRouter / Azure / Grok / Ollama / Custom Model / All Of The Above] working as one.
Run your own MCP server? See who uses it and what to fix.
Measure it with TrackMCP