Nodeon

io.github.taksixPublished by GitHub account taksixv1.0.0Updated Nov 14, 2026

Task graph for coding agents: pull the next unblocked task with scoped context, in dependency order.

VerifiedStreamable HTTPWeb executableProductivity & WorkflowDeveloper Tools

Overview

AI-generated overview

Hosted task-graph server that lets a coding agent pull the next unblocked task, in dependency order, with only the context that task needs.

What you can do
  • List, create, rename and delete projects, and seed one from a starter template
  • Fetch the next unblocked task with its spec, dependencies and dependents, or read the whole graph and build order
  • Mark tasks complete, reopen them, flag blockers and unblock work
  • Add subtasks individually or in batches, edit or delete nodes, and link or unlink dependencies
  • Revoke the assistant's connection to the account
When to use it
Useful when an agentic coding loop drifts, rebuilds work or over-reads context, and you want the plan kept outside the chat as a dependency graph the agent works through one node at a time. Also fits when you want to see and hand-edit the same plan on the web.
Requirements
A remote Streamable HTTP endpoint; nothing runs locally. Sign-in is through OAuth 2.1 on first connection, with no API keys. A Nodeon account is needed, and a free tier is available with paid plans above it.
Before you install
The agent can change your plan: it creates, edits and deletes nodes and projects, adds dependencies, and marks tasks complete, with a force flag that is logged when dependencies are open. Deleting a project or node removes work. The connection can be revoked with the disconnect tool. Paid plans are offered beyond the free tier.

Installation

In SourceWeft

  1. Open Nodeon in the dashboard and add it to a workspace.
  2. Connect the required credentials. They are stored privately and never shown on public pages.
  3. Enable the server for the chats that should use its tools.

Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.

Use with an agent

Paste this prompt into Claude Code, Codex, Cursor or another agent — or into SourceWeft chat. The agent reads this server's install guide, shows you its publisher, endpoint and sign-in, and adds it to your client once you agree. Keys never go into a command.

Read https://sourceweft.com/mcp/io.github.taksix%2Fnodeon/install.md and connect the MCP server it describes to my client. Before changing any config, show me its publisher, endpoint and sign-in method and wait for my OK. Never put an API key in a command or in a file I would commit; ask me where it should come from.

Read the install guide the agent follows

Other MCP clients

Add this to your client's mcpServers config. The client completes OAuth sign-in on first connect.

{
  "mcpServers": {
    "nodeon": {
      "type": "http",
      "url": "https://www.nodeon.app/api/mcp"
    }
  }
}

README

Nodeon MCP

A task graph your coding agent pulls from. Plan your app as a node map, connect your agent, and it builds one node at a time, in dependency order, with only the context that node needs.

This repo holds the public listing, install docs and registry metadata for the hosted Nodeon MCP server. The server itself is hosted; there is nothing to run locally.

Why

Agentic coding loops drift. The agent loses the plot, rebuilds things, and overcomplicates code. Nodeon keeps the plan outside the chat as a dependency graph, so the agent always asks "what's next?" and gets one scoped task back instead of re-reading everything.

  • Next unblocked task with its spec, dependencies and dependents
  • Dependency ordering enforced: a node can't be marked done while its dependencies are open
  • Agent edits the plan too: add subtasks it discovers, flag blockers, unblock, reopen
  • Visual map at nodeon.app to see and edit the same graph by hand
  • Starter templates: SaaS MVP, landing page, mobile app

Install

Paste the URL. Sign-in happens through OAuth the first time your agent connects; no API keys.

Claude Code

bash
claude mcp add --transport http --scope user nodeon https://www.nodeon.app/api/mcp

Cursor (.cursor/mcp.json)

json
{  "mcpServers": {    "nodeon": { "url": "https://www.nodeon.app/api/mcp" }  }}

Codex

bash
codex mcp add nodeon --url https://www.nodeon.app/api/mcpcodex mcp login nodeon

Windsurf (~/.codeium/windsurf/mcp_config.json)

json
{  "mcpServers": {    "nodeon": { "serverUrl": "https://www.nodeon.app/api/mcp" }  }}

Claude.ai / Claude Desktop

Settings → Connectors → Add custom connector → https://www.nodeon.app/api/mcp

Tools

ToolWhat it does
list_projectsList your projects
create_projectStart a new project from chat
apply_templateSeed a project with a starter plan (SaaS MVP, landing page, mobile app)
get_next_taskNext unblocked node with its scoped context
get_project_graphWhole graph: nodes, statuses, edges, build order
get_nodeRe-read one node's spec and links
mark_task_completeMark done; refused while dependencies are open (force flag logged)
add_subtask / batch_add_subtasksAdd discovered work, with dependencies
update_node / delete_nodeEdit or remove nodes
add_dependency / remove_dependencyLink or unlink nodes
flag_blocker / unblock_task / reopen_taskManage blocked and reopened work
rename_project / delete_projectProject management
disconnectRevoke this assistant's connection

Typical loop

  1. list_projects → pick a project
  2. get_next_task → build it
  3. mark_task_complete → repeat
  4. Found new work? add_subtask. Stuck? flag_blocker.

Pricing

Free tier available. Paid plans at https://www.nodeon.app/pricing.

Support

[email protected]

Source: README.md at commit b5276be

Tools

0
This server asks you to sign in, so SourceWeft could not read its tools without an account and has not indexed them. What a connected account sees can depend on that account and its permissions.

Version history

1
  1. v1.0.0LatestNov 14, 2026

Score

75% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 daysRequired25 pts
  • Auth declaredRequired10 pts
  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts

    Declare a license in the repository or the registry entry.