Skip to main content
This guide walks you through your first end-to-end DABAR workflow:
  1. Authenticate against the API.
  2. Create a Politic — the policy that governs how DABAR reasons.
  3. Create a Project — a run that applies the politic to your sources.
  4. Deploy the project so DABAR starts producing verified outputs.
You need a DABAR account and an API token. If you don’t have one yet, sign in at dabarai.com and copy your token from the dashboard.

​
1. Set up your environment

Export your token so the examples below work as-is. 
export DABAR_TOKEN="your_api_token_here"
export DABAR_API="https://api.dabarai.com/v1"

​
2. Verify authentication

List your existing politics to confirm your token works. 
curl "$DABAR_API/politics" \
  -H "api-token: $DABAR_TOKEN"
A 200 response with "success": true means you’re ready. A 401 means the header is missing or the token is invalid — see Authentication.

​
3. Create a Politic

A Politic is a reusable set of rules DABAR must follow when reasoning. It defines what sources are authoritative, what to flag, and how answers should be shaped. 
curl -X POST "$DABAR_API/politics" \
  -H "api-token: $DABAR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Credit Risk Policy",
    "description": "Evaluate loan applications against internal risk guidelines.",
    "rules": [
      { "name": "Scope",    "description": "Only loans above 50,000 USD." },
      { "name": "Evidence", "description": "Every finding must cite source + page." },
      { "name": "Labels",   "description": "Use CONFIRMED / NOT FOUND / ESTIMATED." }
    ],
    "knowledgeFiles": [
      {
        "url": "https://cdn.example.com/risk-guidelines.pdf",
        "type": "pdf",
        "name": "risk-guidelines.pdf"
      }
    ]
  }'
Save the _id returned in data._id — you’ll pass it as politicsId in the next step.

​
4. Create a Project

A Project applies a Politic to a specific set of input files (documents, URLs, audio, video, or raw text). 
curl -X POST "$DABAR_API/projects" \
  -H "api-token: $DABAR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Q1 Loan Batch",
    "description": "Risk analysis for Q1 loan applications.",
    "politicsId": "PASTE_POLITIC_ID_HERE",
    "files": [
      { "name": "application-001.pdf", "type": "document", "url": "https://cdn.example.com/application-001.pdf" },
      { "name": "analyst-notes", "type": "text", "content": "Applicant has prior loan history in 2024..." }
    ]
  }'
The project is created in draft status. It won’t run until you deploy it.

​
5. Deploy the project

curl -X PUT "$DABAR_API/projects/PASTE_PROJECT_ID_HERE" \
  -H "api-token: $DABAR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{ "status": "deploying" }'
DABAR now ingests your files, reasons under the politic’s rules, and produces verified outputs labeled CONFIRMED, NOT FOUND, or ESTIMATED with source and page-level citations. Poll the project with GET /projects/{id} to track progress.

​
Next steps

Policy Engine

Learn how to structure rules that DABAR can enforce end-to-end.

Primary Sources

Understand what kinds of sources DABAR accepts and how they’re used.

Autonomous Agents

Turn a deployed project into an agent that takes action in your systems.

API Reference

Every endpoint, parameter, and response in detail.