Docs › Quickstart

Quickstart

5 minutes to your first Agent Readiness evaluation. No SDK required — just curl.

Prerequisites: You need curl and a terminal. No API key is required for the MVP (see Authentication for details).

Step 1: Create an Evaluation (Quick Eval)

One API call

The quickest way to start is the /api/v1/quick-eval endpoint. It creates a customer, vendors, scenario, and evaluation in a single call.

curl -X POST https://agentreadiness.net/api/v1/quick-eval \
  -H "Content-Type: application/json" \
  -d '{
    "company_name": "YourProduct",
    "website_url": "https://yourproduct.com",
    "competitors": [
      "https://competitor1.com",
      "https://competitor2.com"
    ],
    "task": "Set up email sending via API",
    "category": "email_api"
  }'
Response 
{
  "evaluation_id": "a1b2c3d4-e5f6-7890-abcd-ef1234567890",
  "customer_id": "...",
  "vendor_ids": ["...", "...", "..."],
  "scenario_id": "...",
  "status": "created",
  "message": "Evaluation created. POST /api/v1/evaluations/{id}/run to start."
}

Save the evaluation_id — you will use it in all subsequent steps.

Step 2: Run the Evaluation

Triggers AI agents

Start the agent buying simulation. This triggers multiple LLM models to independently discover, evaluate, and rank the vendors.

curl -X POST https://agentreadiness.net/api/v1/evaluations/{evaluation_id}/run \
  -H "Content-Type: application/json"
Response 
{
  "evaluation_id": "a1b2c3d4-e5f6-7890-abcd-ef1234567890",
  "status": "running",
  "message": "Evaluation started. Check status at GET /api/v1/evaluations/{id}/status"
}

Step 3: Check Status

Poll until complete

The evaluation typically takes 1–3 minutes depending on the number of vendors and models. Poll the status endpoint.

curl https://agentreadiness.net/api/v1/evaluations/{evaluation_id}/status
Response (in progress) 
{
  "evaluation_id": "a1b2c3d4-...",
  "status": "running",
  "progress": {
    "completed_runs": 12,
    "total_runs": 45
  }
}
Response (complete) 
{
  "evaluation_id": "a1b2c3d4-...",
  "status": "completed",
  "progress": {
    "completed_runs": 45,
    "total_runs": 45
  }
}

Step 4: Get Your Scores

Agent Readiness Score

Once complete, retrieve scores across all funnel stages for every vendor and model combination.

curl https://agentreadiness.net/api/v1/evaluations/{evaluation_id}/scores
Response 
{
  "evaluation_id": "a1b2c3d4-...",
  "overall_score": 72.5,
  "funnel_scores": {
    "consideration": 95.0,
    "comprehension": 80.0,
    "preference": 65.0,
    "executability": 55.0,
    "completion": 40.0
  },
  "vendor_rankings": [...]
}

Step 5: Get Diagnosis and Report

Actionable fixes

The diagnosis endpoint returns root-cause analysis and engineering-ready recommendations.

# Get causal diagnosis
curl https://agentreadiness.net/api/v1/evaluations/{evaluation_id}/diagnosis

# Get the full HTML report
curl https://agentreadiness.net/api/v1/evaluations/{evaluation_id}/report

The HTML report includes funnel visualizations, competitive comparison, and prioritized recommendations.

Done! You now have a complete Agent Readiness evaluation. Share the report with your team and start fixing the issues identified in the diagnosis.

Next Steps