Claude Code

Claude Code has two One Horizon paths. A local Claude Code agent worker runs queued implementation work from your machine. The MCP connector lets you ask Claude Code about One Horizon work while you stay in the terminal.

For Claude Desktop or Claude Web chat instead of terminal coding, use Claude. Claude is the Desktop & Web MCP connector path; Claude Code is the terminal and local-agent path.

Local agent setup is the important path for delegated work. It uses the authenticated claude CLI on your machine, not Claude Desktop and not the MCP connector configuration.

Choose the Claude Code path

Path	Best for	What happens
Local Claude Code agent	Queued implementation, planning, review, and workflow steps.	One Horizon starts the claude CLI on your machine, sends the work item as the brief, and records the run.
MCP connector	Asking Claude Code about workspace work from an interactive terminal chat.	Claude Code calls the hosted One Horizon MCP server with your authorized user access.
Open in Claude Code	Resuming a stopped local agent run.	Completed or failed runs can reopen the saved Claude Code session from the run menu.

MCP setup does not satisfy local agent prerequisites. The local agent needs the claude binary on PATH and an authenticated Claude Code CLI session.

What the Claude Code worker can do

A local Claude Code agent worker can:

• Claim One Horizon sessions from Send to, direct assignment, or workflow steps.
• Run with Builder, Planner, or Reviewer presets.
• Use Read-only, Workspace access, or Full access according to the agent configuration.
• Prepare the task workspace before Claude Code starts.
• Follow the workflow file for setup, checks, commits, and handoff rules.
• Report progress, terminal activity, provider run metadata, and final outcome back to the run.
• Reopen a completed or failed Claude Code session when a saved provider session ID exists.

One Horizon does not collect Claude credentials. Claude Code authentication remains owned by the Claude Code CLI on this machine.

Set up a Claude Code local agent

1. 1

   Install and sign in to Claude Code CLI

   Confirm the claude command is available in the same environment that starts Desktop agents or runs one agent:

   claude --versionclaude /login

   Claude Desktop is not the automation runtime. The local agent invokes the CLI directly.

2. 2

   Install Desktop or sign in to the CLI

   Use One Horizon Desktop for guided setup, or use the One Horizon CLI from the repository.

3. 3

   Create the Claude Code agent

   In Desktop, open Workflows, go to the agents page, select Add agent, and choose Claude Code. In the terminal, run:

   one agent create

   Choose the runtime preset, agent access, project folder, code isolation, workflow file, and PR or MR behavior.

4. 4

   Verify the runtime

   In Desktop, finish setup and confirm the agent shows online. In the terminal, run:

   one agent doctor

   Fix any missing claude binary, CLI authentication, workspace, repository, or workflow file issue before starting pickup.

5. 5

   Run assigned work

   Open an initiative, bug, or todo with a usable brief and choose Send to, or target the agent from a workflow step. For terminal pickup, run:

   one agent watch

Use Local Agents for runtime presets, access levels, project folders, worktrees, pull requests, and restart behavior. Use Monitoring for run states, recovery actions, and paused sessions.

Prepare the work item

Claude Code works best when the initiative, bug, or todo already explains the expected change. Add the outcome, important files or links, constraints, and acceptance notes before launch.

For planned implementation work, keep the brief close to spec-driven development: specific enough for Claude Code to act, short enough for a person to review.

Connect Claude Code through MCP

MCP is the interactive path. Configure it when you want Claude Code chat to read or update One Horizon work while you code.

The Claude Code integration uses the hosted One Horizon MCP server over HTTP with OAuth.

1. 1

   Add the MCP server

   Import an existing Claude Desktop MCP configuration:

   claude mcp add-from-claude-desktop --scope user

   Copied!Copy to clipboard

   Or add One Horizon directly:

   claude mcp add --transport http onehorizon https://mcp.onehorizon.ai/mcp

   Copied!Copy to clipboard

2. 2

   Authorize the connection

   Run /mcp in Claude Code, open the browser authorization flow, sign in to One Horizon, and return to Claude Code.

   /mcp

   Copied!Copy to clipboard

   Use Clear authentication in the /mcp menu when you need to revoke and reconnect.

3. 3

   Share project configuration when needed

   Commit .mcp.json at the repository root when the project should carry the One Horizon MCP entry for the team.

   {  "mcpServers": {    "onehorizon": {      "type": "http",      "url": "https://mcp.onehorizon.ai/mcp"    }  }}

   Copied!Copy to clipboard

   The hosted MCP server uses OAuth, so this file does not need environment variables.

Ask Claude Code through MCP

Short prompts are enough when the work item already has detail:

"show my assigned issues and todos for today"

Claude Code can pull assigned work from One Horizon, including work connected from Linear or Jira.

"use my assigned work details for the gateway API"

Claude Code can bring the work item into context before proposing changes.

"capture this as completed work"

Claude Code can summarize the result and add a completed task.

Security and permissions

The MCP connector acts on your behalf. Data access follows your roles, workspace, team membership, and granted scopes.

Local agent runs use the agent access level you chose during setup. The selected access level is the boundary for commands, files, and provider tools.

FAQs

Local agent FAQs

MCP FAQs