blinkdesk

Blink Desk ==========

A lightweight ticketing system for agent orchestration, developers, and command-line workflows. Uses SQLite with Python stdlib only.

---

Quick Start ==========

Python Library --------------

pip install blinkdesk

from blinkdesk import TicketingSystem

system = TicketingSystem("tickets.db")
system.init_database()

ticket = system.create_ticket(
    title="Initial setup complete",
    description="Database initialized and ready for use"
)
print(f"Created ticket #{ticket.id}: {ticket.title}")

MCP Server ----------

pip install blinkdesk[mcp]
bd -d tickets.db init
bd -d tickets.db mcp stdio

Add to your agent config:

{
  "mcpServers": {
    "blinkdesk": {
      "command": "bd",
      "args": ["-d", "/path/to/tickets.db", "mcp", "stdio"]
    }
  }
}

HTTP and SSE transports also supported.

CLI ---

bd -d tickets.db ticket create --title "Test"
bd -d tickets.db ticket list

---

Core Concepts ============

Tickets track issues, TODO items, bugs, anything - they represent work that needs to be tracked. Each ticket has a title, optional description, a current state, and optionally an assignee.

Entities are the people, teams, or agents who can own tickets. Think of them as your users or groups. Each entity has a slug (like "alice" or "support-team") and a display name.

States are the stages tickets can be in - like TODO, IN_PROGRESS, and DONE. You define these in your schema file, so the workflow matches how your team actually works.

Comments let you add discussion to a ticket. Useful for context, updates, or just keeping a record of decisions.

Audit Log records every change - who did what and when. Great for accountability and debugging.

---

Features ========

• Zero dependencies - Pure Python with SQLite
• MCP server - Full AI assistant integration
• Python API - Clean, typed interface
• CLI interface - Scriptable command-line tools
• **Audi