mcpField
U
Unblocked

Unblocked

Remote
getunblocked.comMemory & Knowledge16Updated 2d ago

About Unblocked

Unblocked provides coding agents with organizational context from code, pull requests, documentation, issues, and team conversations. It is aimed at development teams and pairs a hosted MCP server for research and retrieval with agent skills and CLI connection guidance.

Key features

  1. 01Research context across code and team sources
  2. 02Retrieve content from known URLs
  3. 03Get repository coding rules
  4. 04Provide skills for agents
  5. 05Include CLI and client setup guidance

Use cases

  1. 01Research code, pull requests, and related discussions
  2. 02Read content from documentation or issue URLs
  3. 03Retrieve repository conventions before coding
  4. 04Configure organizational context for coding agents

MCP support

The hosted MCP server connects over Streamable HTTP with OAuth and exposes `context_research`, `context_get_urls`, and `context_get_rules`. This repository provides connection documentation and skills, not the hosted server implementation.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://getunblocked.com/api/mcpsse
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

Details

From the project README

Unblocked MCP and agent skills

Unblocked gives coding agents organizational context from code, pull requests, docs, issues, and conversations. Unblocked MCP makes that context available through the Model Context Protocol. Agent skills teach agents when and how to use it.

This repository contains public connection documentation, MCP Registry metadata, and agent skills. It does not contain the hosted server's implementation. The skills follow the open Agent Skills specification.

Connect to hosted Unblocked MCP

You need an Unblocked account with access to your team's workspace. To set up Unblocked, first connect your source code. Then add data sources such as documentation, issue trackers, and messaging platforms to give your agents context beyond the code. Follow the MCP setup guide to connect your coding agent.

  1. Add a remote MCP server in a client that supports Streamable HTTP and OAuth.
  2. Set the server URL to https://getunblocked.com/api/mcpsse.
  3. Select Streamable HTTP, sometimes shown as HTTP in the client.
  4. Follow the client's OAuth prompt. Sign in to Unblocked and authorize the connection.

See the current remote connection instructions for client setup and authentication details.

MCP tools

The hosted MCP server exposes three tools:

MCP toolPurpose
context_researchResearch across code, pull requests, docs, issues, and conversations. Returns an answer with source citations.
context_get_urlsRetrieve content from one or more known URLs.
context_get_rulesRetrieve a repository's coding rules and conventions.

MCP tools, CLI commands, and skills are separate. The Unblocked CLI provides commands such as unblocked context-research and unblocked context-get-urls. It also provides context-search-* and context-query-* commands. Those extra commands are not hosted MCP tools. The skills below guide agents through these workflows.

Available skills

SkillDescription
unblocked-context-researchResearch across connected sources using MCP or the CLI
unblocked-context-search-codeSearch code across connected repositories using the CLI
unblocked-context-search-prsSearch pull request descriptions and reviews using the CLI
unblocked-context-search-issuesSearch issue trackers using the CLI
unblocked-context-search-documentationSearch documentation using the CLI
unblocked-context-search-messagesSearch team conversations using the CLI
unblocked-context-query-issuesRetrieve issues filtered by project and person using the CLI
unblocked-context-query-prsRetrieve pull requests filtered by repository and person using the CLI
unblocked-context-get-urlsRetrieve URL content using MCP or the CLI
unblocked-context-get-rulesRetrieve a repository's coding rules using MCP or the CLI
unblocked-tools-guideChoose between CLI commands and MCP tools, run headless, and handle unavailable tools

Setup

Unblocked CLI installer

On macOS and Linux, this command installs the Unblocked CLI, configures supported coding agents, installs skills, and starts OAuth authentication:

curl -fsSL https://getunblocked.com/install-mcp.sh | bash

See the MCP setup guide for supported clients and platform instructions.

Claude plugin

Connect this repository as a plugin directly in Claude Code:

claude plugin marketplace add unblocked/skills
claude plugin install unblocked-skills
Devin plugin

Install the plugin for your personal Devin scope:

devin plugins install unblocked/skills

To install it for an organization or enterprise, open Customize, select Add plugin, choose From repository, and enter unblocked/skills.

Cursor plugin

This repository is a single Cursor plugin. It installs all skills and configures the hosted Unblocked MCP server.

To test it as a local plugin:

mkdir -p ~/.cursor/plugins/local
git clone https://github.com/unblocked/skills ~/.cursor/plugins/local/unblocked-skills

Restart Cursor or run Developer: Reload Window. Open Customize and confirm that the Unblocked skills and MCP server are present. Cursor will start the Unblocked OAuth flow when the MCP server needs authentication.

Using the Skills CLI
bunx skills add unblocked/skills
Manual copy

Copy the skill directory into your agent's skills folder. Replace <skill-name> with a directory such as unblocked-context-research or unblocked-context-get-urls:

# Claude Code
cp -r skills/<skill-name> ~/.claude/skills/

# Cursor
cp -r skills/<skill-name> .cursor/skills/

Documentation and support

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.