nerdocs

Ntasker — Development skill for Claude Code

Development community

local tasks helper for Claude (and other AIs).

How to install Ntasker

This entry records only its repository, not the path inside it, so there is no exact command to give. Open nerdocs/ntasker and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Ntasker does

local tasks helper for Claude (and other AIs).

Alternatives in Development

Ralph Loop — Autonomous coding loop — works through tasks, commits, repeats 14k ★
Om Meeting — Prep for any meeting by topic — subject-forward briefing with open tasks, blockers, and brainstormed considera 4.6k ★
Code Review Graph — Local knowledge graph for Claude Code 3.7k ★

README

ntasker

Lightweight local task tracker. Single-user, FastAPI + SQLite, Tabler.io UI.

[ Buy me a coffee ](https://buymeacoffee.com/nerdoc)

In your AI coding agent

ntasker is **agent-agnostic**: every task can run on **Claude Code, OpenCode or Pi** (and the framework is extensible to more). It doubles as your agent's task memory -- the [skill + `/task` command](#ai-agent-integration) let the agent read and drive your tracker, no copy-paste:

"What should I work on next?" -- the agent grabs the open tasks for your current project folder and ranks them by urgency.
/task 34 -- pulls #34 into the session (title, description, tags), flips it to in progress, and warns you if you're sitting in the wrong project.
"Add a todo: ..." -- it files the task for you; drop a #34 anywhere later and it knows exactly which task you mean.
Finished an assigned task? The agent moves it to Review for you to sign off -- it never closes, deletes, or archives tasks on its own.

![ntasker kanban board with the projects sidebar](docs/screenshot.jpg)

Run with an agent (web UI)

The flip side of the integration above: every task row has a **run** button -- showing that task's agent logo -- that opens a real interactive session -- the genuine TUI, embedded in the page via xterm.js -- running in the task's project directory and seeded with `/task `. You answer the agent's questions, approve its tool prompts and interrupt it exactly as in a terminal; it is the same CLI with the same `CLAUDE.md`, skills, MCP and permissions.

Each task picks its agent (or inherits the `default_agent` setting); the run button only appears when that agent's CLI resolves. Sessions run in the background (the button shows a spinner, and re-opening reattaches to the live session); marking a task **done** ends its session. Needs the agent's CLI on `PATH` (or a configured path) and a PO