Skip to main content
Claude Code is Anthropic’s official CLI tool. This guide shows how to configure Claude Code with context about COR APIs for building integrations.

​
Prerequisites

  • Active Claude subscription (Pro, Max, or API access)
  • COR API credentials (API Key and Client Secret)

​
Setup

  1. Install Claude Code globally:
npm install -g @anthropic-ai/claude-code
  1. Navigate to your project directory.
  2. Add the CLAUDE.md file below to your project root.
  3. Run claude to start.

​
Create CLAUDE.md

Create a CLAUDE.md file at the root of your project to provide Claude Code with COR API context: 
# COR API Integration

## Working Relationship

- Ask for clarification rather than making assumptions about API usage
- Verify endpoint paths and required fields before generating code
- Reference the COR API documentation when uncertain

## COR Platform

COR is a project management platform for agencies. Three APIs are available:

| API | Base URL |
|-----|----------|
| Main API | `https://api.projectcor.com/v1` |
| Resource Allocation | `https://planner.svc.v2.projectcor.com` |
| Integrations | `https://integrations.projectcor.com` |

## Authentication

OAuth 2.0 with Bearer tokens (same token for all APIs).

```bash
# Get token
curl -X POST 'https://api.projectcor.com/v1/oauth/token?grant_type=client_credentials' \
  -H 'Authorization: Basic BASE64(API_KEY:CLIENT_SECRET)'

# Use token
Authorization: Bearer ACCESS_TOKEN
```

## Main API Entities

- **Clients** `/clients` - Customer accounts
- **Projects** `/projects` - Project management
- **Tasks** `/tasks` - Task tracking
- **Hours** `/hours` - Time entries
- **Users** `/users` - Team members
- **Brands** `/brands` - Client brands
- **Products** `/products` - Brand products

### User Roles

`1` C-Level, `2` Director, `3` Project Manager, `4` Collaborator, `5` Freelancer, `6` Client

### Task Status

`nueva` (new), `en_proceso` (in progress), `estancada` (stalled), `finalizada` (completed)

### Project Status

`in_process`, `finished`, `suspended`

## Data Conventions

### Pagination

```json
{"total": "45", "perPage": 20, "page": 1, "lastPage": 3, "data": [...]}
```

- `page`: Page number (default 1, set `false` for all)
- `perPage`: Items per page (default 20)

### Filters

URL-encoded JSON: `?filters=%7B%22client_id%22%3A123%7D`

### Dates

Request: `YYYY-MM-DD` | Response: `YYYY-MM-DD HH:mm:ss`

## Resource Allocation API

Manages user capacity on projects.

### Endpoints

- `POST /allocation/saveAllocation` - Create
- `PUT /allocation/updateAllocation` - Update
- `DELETE /allocation/removeAllocation` - Delete
- `GET /allocation/getAllocations/{projectId}` - List

### Allocation Object

```json
{
  "userId": 1111,
  "companyId": 1111,
  "resource": {"typeId": "Project", "entityId": 1111},
  "from": "2024-06-01",
  "to": "2024-10-31",
  "assignedByUserId": 11111,
  "allocatedHours": 20
}
```

### Validation

- `from` < `to`, max 12 months range
- No overlapping allocations
- `allocatedHours` required

## Integrations API

Syncs external systems (Salesforce, Jira, SAP, Okta, etc.) with COR.

### Required: metadata.source

```json
{
  "metadata": {"source": "SALESFORCE"},
  "id": "external-id",
  "name": "Entity Name"
}
```

### Sources

`JIRA`, `SALESFORCE`, `ADVERTMIND`, `QUICKBOOKS`, `ZAPIER`, `OKTA`, `MICROSOFT_DYNAMICS`, `GITHUB`, `MICROSOFT_TEAMS`, `VBS`, `SAP`, `GLOBANT`

### Entities

Projects, Clients, Users, Working Time, Contracts, Workspaces, Brands

### Error Codes

`ZC001` Generic, `ZC002` Validation, `ZC003` API Request, `ZC004` Query, `ZC005` External Request

## HTTP Status Codes

`200` OK, `201` Created, `400` Bad Request, `401` Unauthorized, `403` Forbidden, `404` Not Found, `422` Validation Error, `429` Rate Limited, `500` Server Error

## Do Not

- Hardcode API credentials in source code
- Skip token refresh handling
- Ignore pagination on list endpoints
- Omit `metadata.source` on Integration API calls
- Send invalid date formats (use YYYY-MM-DD)
- Create overlapping resource allocations