Getting Started with Rekor

Rekor is a headless system of record built for AI agents. Agents create collections, upsert records, and link relationships via MCP tools, CLI, or REST API.

Core Concepts

Collections

A collection defines a type of record using JSON Schema. Created at runtime — no migrations needed.

Records

JSON documents that conform to a collection's schema. Upsert by external ID for idempotent writes from external systems.

Relationships

Typed, directed links between two records with optional metadata. First-class entities — queryable in any direction.

Quick Start

1. Install the CLI

npm install -g @rekor/cli
rekor login

2. Create a workspace

rekor workspaces create my-workspace --name "My Workspace"

3. Create a collection

rekor collections upsert contacts \
  --workspace my-workspace \
  --schema '{"type":"object","required":["name"],"properties":{"name":{"type":"string"},"email":{"type":"string"}}}'

4. Add a record

rekor records upsert contacts \
  --workspace my-workspace \
  --external-id "c_001" --source crm \
  --data '{"name":"Jane Doe","email":"jane@example.com"}'

5. Query records

rekor query contacts \
  --workspace my-workspace \
  --filter '{"field":"data.name","op":"like","value":"%Jane%"}'

Interfaces

Interface	Best For
MCP Tools	AI agents (Claude, ChatGPT, etc.)
CLI	Scripts, automation, human operators
REST API	Integrations, webhooks, custom apps

Next Steps

• Skill — load Rekor's skill into any AI agent so it can use the CLI as its data layer.
• Environments — understand preview and production workspaces before deploying.
• Integrations — connect external systems via hooks, triggers, batch operations, and provider adapters.