Skip to main content
Glama

Redmine MCP Server

An MCP server that lets a coding agent read and write your Redmine: search tickets, read full issue histories, post notes, log time, browse the wiki, and download attachments. Works with Claude Code, Codex, Claude Desktop, and any other MCP client.

It ships with a built-in safety rule: the agent will never create, update, comment on, or log time against a ticket unless you explicitly tell it to in that conversation.


Setup

Step 1: Install uv

uv runs the server and handles its Python dependencies for you.

curl -LsSf https://astral.sh/uv/install.sh | sh

On Windows, use powershell -c "irm https://astral.sh/uv/install.ps1 | iex".

Step 2: Download this server

git clone https://github.com/beborico1/mcp-redmine.git
cd mcp-redmine
uv sync

Note the full path you cloned into. You need it in step 4:

pwd

Step 3: Get your Redmine API key

  1. Log in to your Redmine.

  2. Go to My account (top right).

  3. In the right-hand sidebar, find API access key and click Show.

  4. Copy the key.

If you do not see that panel, your administrator has not enabled the REST API. Ask them to tick Enable REST web service under Administration → Settings → API.

Step 4: Connect it to your agent

Pick the one you use. In every case, substitute your Redmine address, your API key, and the path from step 2.

Claude Code

claude mcp add redmine \
  --env REDMINE_URL=https://redmine.example.com \
  --env REDMINE_API_KEY=paste_your_key_here \
  -- uv --directory /full/path/to/mcp-redmine run mcp-redmine

Codex

Open ~/.codex/config.toml (create it if it does not exist) and add:

[mcp_servers.redmine]
command = "uv"
args = ["--directory", "/full/path/to/mcp-redmine", "run", "mcp-redmine"]

[mcp_servers.redmine.env]
REDMINE_URL = "https://redmine.example.com"
REDMINE_API_KEY = "paste_your_key_here"

Claude Desktop, or any other MCP client

Add this to the client's MCP config file:

{
  "mcpServers": {
    "redmine": {
      "command": "uv",
      "args": ["--directory", "/full/path/to/mcp-redmine", "run", "mcp-redmine"],
      "env": {
        "REDMINE_URL": "https://redmine.example.com",
        "REDMINE_API_KEY": "paste_your_key_here"
      }
    }
  }
}

Step 5: Check that it works

Restart the agent, then ask it:

What's on my Redmine dashboard?

You should get your open tickets back. That is it, you are done.


Related MCP server: Redmineflux MCP Server

Configuration

Variable

Required

Description

REDMINE_URL

yes

Base URL of your Redmine, no trailing slash

REDMINE_API_KEY

yes

Your personal API access key from My account

Your API key is read from the environment only. It is never written to disk by this server, and it must never be committed to a repository.


What you can ask for

Reading

  • redmine_my_dashboard - your open tickets, at a glance

  • redmine_get_issue - one ticket in full, with its comment history

  • redmine_search_issues - filter by author, assignee, project, status, dates

  • redmine_list_projects - every project you can see

  • redmine_list_wiki_pages, redmine_get_wiki_page - project wikis

  • redmine_get_time_entries - logged time

  • redmine_download_attachment - save a ticket's files locally

  • redmine_find_user, redmine_get_users - resolve a name to a user ID

  • redmine_get_statuses, redmine_get_trackers, redmine_get_priorities - reference IDs

Writing (each one needs your explicit go-ahead)

  • redmine_create_issue - open a new ticket, attachments included

  • redmine_update_issue - change status, assignee, or any field

  • redmine_add_note - post a comment

  • redmine_edit_journal - edit a comment you already posted

  • redmine_log_time - record hours against a ticket

Ask in plain language. "Show me ticket 12345", "find everything Yuko opened this month", "draft a status update for 12345" all work; the agent picks the tool.


Troubleshooting

Every call fails with a 401. The API key is wrong, or REDMINE_API_KEY did not reach the server. Re-copy it from My account and re-run the claude mcp add command.

Every call fails with a 404. Check REDMINE_URL. It should be the site root (https://redmine.example.com), not a project or issue page, and it should have no trailing slash.

The server shows as failed with -32000: Connection closed. The server process is exiting before it can talk to the client. Run the command from your config by hand to see the real error:

uv --directory /full/path/to/mcp-redmine run mcp-redmine

Failed to spawn: mcp-redmine means step 2 was skipped or did not finish, so run uv sync in the clone. command not found: uv means uv is not on the PATH your client inherits, so use its absolute path (which uv) in the config.

The tools are not available. Restart the client after adding the server, then confirm it is connected with claude mcp list or codex mcp list.

Searching by a person's name returns a 500. Redmine only accepts me or a numeric user ID in author and assignee filters. Ask the agent to look the person up with redmine_find_user first.


Requirements

Python 3.13 or newer, and a Redmine with the REST API enabled.

Available Tools

18 tools
redmine_add_noteA

Add a comment/note to an existing Redmine issue. This is the primary way to post updates, progress reports, investigation findings, or review feedback on a ticket. Follow the communication style guide: structured formatting, evidence-based, warm closing.

Inputs: notes (required), issue_id (required), status_id, done_ratio, +2 more.

redmine_create_issueA

Create a new Redmine issue. The description should be well-structured: brief statement of what needs to be done, context/why, scope, and links. For bugs: include steps to reproduce, environment, and screenshots.

Inputs: subject (required), project_id (required), description (required), due_date, +12 more.

redmine_download_attachmentA

Download a file attachment from Redmine to the local filesystem. Use redmine_get_issue first to see available attachments and their IDs. Returns the local file path so you can read or process the file.

Inputs: attachment_id (required), save_dir.

redmine_edit_journalA

Edit the text of an existing journal (comment/note) on a Redmine issue. Use redmine_get_issue first to find the journal ID you want to edit. Only the notes text can be changed. Requires Redmine 5.0+.

Inputs: notes (required), journal_id (required).

redmine_find_userA

Find a Redmine user ID by scanning recent issues for matching authors and assignees. Use this as a workaround for /users.json admin restrictions: given a name fragment (e.g. 'Toyoda', 'yoshikuni', 'Murata'), it returns matching user IDs sorted by how often they appear, so the most active match comes first. Once you have an ID, pass it to redmine_search_issues as author/assigned_to/updated_by.

Inputs: name (required), scan_pages.

redmine_get_issueA

Get full details of a Redmine issue including description, metadata, and (by default) all journal updates and attachments. Set include_journals=False for tickets with very long histories that would blow up your context (e.g. recurring incident tickets). Use this to understand the full history of a ticket before updating it.

Inputs: issue_id (required), include_children, include_journals, include_relations, +1 more.

redmine_get_prioritiesA

List all available issue priorities with their IDs.

Inputs: no named parameters.

redmine_get_statusesA

List all available issue statuses with their IDs. Use this to find the right status_id for updates.

Inputs: no named parameters.

redmine_get_time_entriesB

Get time entries for a user or issue, useful for tracking and reporting.

Inputs: limit, to_date, user_id, issue_id, +1 more.

redmine_get_trackersA

List all available issue trackers with their IDs. Use this to find the right tracker_id.

Inputs: no named parameters.

redmine_get_usersA

Search for Redmine users by name. Useful for finding user IDs when assigning issues or @mentioning colleagues.

Inputs: name, limit.

redmine_get_wiki_pageA

Get the full content of a specific wiki page in a Redmine project. Returns page metadata (author, dates, parent page) and the raw page content (this Redmine instance uses standard Markdown, not textile). Use redmine_list_wiki_pages first to find page titles.

Inputs: page_title (required), project_id (required), version.

redmine_list_projectsA

List all Redmine projects you have access to, with their identifiers and status.

Inputs: limit.

redmine_list_wiki_pagesA

List all wiki pages in a Redmine project. Returns a hierarchical tree showing page titles, versions, and last-updated dates. Use this to discover available documentation before reading specific pages.

Inputs: project_id (required).

redmine_log_timeB

Log time spent on an issue. Include a meaningful comment about what was accomplished during this time.

Inputs: hours (required), issue_id (required), comments, spent_on, +1 more.

redmine_my_dashboardA

Get a quick overview of your Redmine workload: issues assigned to you, issues you recently updated, and overdue items. Use this as the first tool when starting a work session to see what needs attention.

Inputs: include_closed.

redmine_search_issuesA

Search Redmine issues with flexible filters. Filter by project, status, author, assignee, tracker, priority, free-text, and date ranges (created/updated). Supports offset for pagination through total_count. Returns a summary list -- use redmine_get_issue for full details. NOTE: author/assigned_to/updated_by accept 'me' or a numeric user ID -- display names like 'Yuko Matsui' will fail (Redmine API returns 500). Use redmine_find_user to resolve a name to an ID first.

Inputs: sort, limit, query, author, +11 more.

redmine_update_issueA

Update issue fields (status, assignee, priority, due date, etc.) without necessarily adding a note. Use redmine_add_note if you want to post a comment. Use this for pure metadata changes.

Inputs: issue_id (required), notes, subject, due_date, +5 more.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.6/5.0

    Scored across 18 tools

    Disambiguation3/5

    Most tools have distinct purposes, but redmine_get_users and redmine_find_user overlap in resolving user IDs, and redmine_my_dashboard partially duplicates redmine_search_issues for personal queries. These overlaps create mild selection ambiguity.

    Naming Consistency4/5

    All tools share the redmine_ prefix and nearly all follow a verb_noun pattern (get_issue, create_issue, list_projects). The exception is redmine_my_dashboard, which uses a possessive instead of a verb, breaking the pattern slightly.

    Tool Count3/5

    With 18 tools, the server falls in the borderline heavy range (16-25). The count is justified by the broad domain coverage (issues, users, time, wiki, projects), but it feels slightly oversized for a focused MCP.

    Completeness4/5

    The core issue workflow is well covered: create, retrieve, update, add notes, search, and log time are all present. However, missing delete operations for issues or notes, no attachment upload, and read-only wiki access represent minor but notable gaps.

    Maintenance

    ActivityInactive
    ResponsivenessLimited data

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      B
      quality
      D
      maintenance
      MCP server for Redmine project management, enabling tools for managing projects, issues, users, time entries, groups, memberships, versions, wiki, news, attachments, search, and Agile sprints via the Redmine REST API.
      89
      21 npm
      1
      MIT
    • A
      license
      Not graded
      quality
      D
      maintenance
      Provides an MCP server for interacting with Redmine's REST API, enabling LLMs to manage issues, projects, users, and time entries.
      MIT

    Appeared in Searches