# claudia

Task management system for AI agents featuring hierarchical subtasks, sprint tracking, prerequisite checks, and 64KB context per issue.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 15
- Author: yuvalsuede
- License: GPL-3.0 · policy=link-only · description only; the file stays with the author
- Source: https://github.com/yuvalsuede/claudia
- Updated: 2026-02-28
- HTML: https://affpapa.org/en/nevibecode/mcp/claudia

## What it does

claudia provides an operational task tracking layer that assistants manage right inside their workspace. Agents break down complex requests into nested subtasks and verify explicit acceptance criteria before marking an item finished.

The runtime enforces a strict transition pipeline from `pending` to `completed`, prevents cyclic dependencies, and assigns 64KB of JSON memory to every task. Multi-agent workflows stay aligned through atomic task claiming and optimistic locks. A built-in kanban board is also included on port `3333`.

Building the standalone binary from source requires Bun runtime version 1.0 or newer.

## Requirements
- Needs: Node.js (commands run via npx)

## Install

### Claude Code (native platform)

**Platform CLI**
```bash
claude mcp add claudia -- npx -y claudia-cli
```

### Codex

**Platform CLI**
```bash
codex mcp add claudia -- npx -y claudia-cli
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "claudia": {
            "command": "npx",
            "args": [
                "-y",
                "claudia-cli"
            ]
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "claudia": {
            "type": "stdio",
            "command": "npx",
            "args": [
                "-y",
                "claudia-cli"
            ]
        }
    }
}
```

### Gemini CLI

**Platform CLI**
```bash
gemini mcp add claudia npx -y claudia-cli
```

## When to use it
- structure complex assignments into nested subtask hierarchies
- prevent execution of blocked issues until prerequisites finish
- persist up to 64KB of JSON metadata across model sessions
- coordinate concurrent agents with atomic claims and transfers

## Example prompts
- Create a task titled "Implement login" with acceptance criteria "Validate password length"
- List ready tasks whose blocking dependencies are satisfied
- Display the full task tree for the active project

## Similar
- [KanbanThing](https://affpapa.org/en/nevibecode/mcp/kanbanthing.md) — MCP server · ★ 11
- [game-of-cards](https://affpapa.org/en/nevibecode/plugins/game-of-cards.md) — plugin · ★ 7
- [kanban-ai](https://affpapa.org/en/nevibecode/skills/kanban-ai.md) — skill · ★ 25
- [plan-harder](https://affpapa.org/en/nevibecode/skills/plan-harder.md) — skill · 1200 installs
- [planner](https://affpapa.org/en/nevibecode/skills/planner.md) — skill · 1200 installs
- [jira-cli](https://affpapa.org/en/nevibecode/skills/jira-cli.md) — skill · 635 installs

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md