Docs

Prompt engineering with metrics

Turn prompt engineering into measurable work. Create a project, copy the key, and send one call.

ProjectsCreate project
Current step
Create a project
Next step
Copy the project key
1Create a project

Start with one project for one app or one wrapper.

2Copy the key

Use the project key in your code or SDK setup.

3Send one call

Send prompt, output, and modelName to /api/ingest.

Setup in order
  1. 1. Create a project for one app or one wrapper.
  2. 2. Copy the project API key.
  3. 3. Send `prompt`, `output`, and `modelName` to `/api/ingest`.
  4. 4. We generate metrics and measure the call.
That is the product. No setup maze. No extra SDK required.
track.ts
await fetch("https://api.agentplasticity.com/api/ingest", {
  method: "POST",
  headers: {
    "Content-Type": "application/json",
    Authorization: "Bearer " + token,
  },
  body: JSON.stringify({
    prompt,
    output,
    modelName: "gpt-5.5",
  }),
});
What we store
Prompt and model output
Metric definitions and scores
Latency and token usage
Evaluation history over time