atlassian-server-mcp
A TypeScript MCP server that enables LLM agents to interact with Atlassian Data Center products (Jira, Confluence, Bitbucket) through typed tools, with tiered operation exposure and safety defaults.
README
<div align="center"> <img src="https://raw.githubusercontent.com/zaimokuza-yoshiteru/atlassian-server-mcp/main/assets/banner.svg" alt="atlassian-server-mcp" width="960"> <br> <a href="https://www.npmjs.com/package/@zaimokuza/atlassian-server-mcp"><img src="https://img.shields.io/npm/v/@zaimokuza/atlassian-server-mcp" alt="npm version"></a> <a href="https://github.com/zaimokuza-yoshiteru/atlassian-server-mcp/actions/workflows/ci.yml"><img src="https://img.shields.io/github/actions/workflow/status/zaimokuza-yoshiteru/atlassian-server-mcp/ci.yml?branch=main" alt="CI status"></a> <a href="https://github.com/zaimokuza-yoshiteru/atlassian-server-mcp/blob/main/LICENSE"><img src="https://img.shields.io/github/license/zaimokuza-yoshiteru/atlassian-server-mcp" alt="license: MIT"></a> <a href="https://github.com/zaimokuza-yoshiteru/atlassian-server-mcp/blob/main/package.json"><img src="https://img.shields.io/badge/node-%3E%3D22-339933?logo=node.js&logoColor=white" alt="node >= 22"></a> </div>
Atlassian Server MCP
A single TypeScript MCP 2.x stdio server for Jira, Confluence, and Bitbucket
Data Center. It combines high-frequency typed tools with a fixed REST
operation registry discovered through generic tools, so the complete registry
never floods tools/list.
Features
- Three products, one server — Jira DC platform APIs plus the Jira Software Agile APIs (boards, sprints, backlog, epics), Confluence, and Bitbucket pull-request workflows.
- Tiered exposure — operations carry an intrinsic
read/safe/risky/maxtier; the agent only sees what your configured tier and FORCE globs allow. An agent exposure boundary, not an authorization boundary. - Safe by default — starts read-only, TLS verification on, PAT auth preferred, credentials redacted from every error and log line.
- Policy generated from official specs — the exposure policy is built
from pinned Atlassian OpenAPI snapshots;
atlassian_discover_operationsreports exactly what a running server exposes. - Strict pre-flight
doctor— validates configuration, credentials, reachability, TLS, and file-root writability before you wire up a client.
Quick start (5 minutes)
-
Install and build (Node.js 22 or newer):
pnpm install pnpm build -
Configure at least one product. A product is enabled when its URL is present; a non-empty product PAT wins, otherwise the shared Basic credentials are required:
export JIRA_URL=https://jira.example.internal export JIRA_TOKEN=... # optional: CONFLUENCE_URL / CONFLUENCE_TOKEN, BITBUCKET_URL / BITBUCKET_TOKEN # Basic fallback (shared): ATLASSIAN_USERNAME + ATLASSIAN_PASSWORD -
Point your MCP client at the server:
{ "mcpServers": { "atlassian-dc": { "command": "node", "args": ["/absolute/path/to/atlassian-server-mcp/dist/cli.js"], "env": { "JIRA_URL": "https://jira.example.internal", "JIRA_TOKEN": "..." } } } }When the package is published,
npx @zaimokuza/atlassian-server-mcpworks the same way. -
Verify with the strict pre-flight check:
npx @zaimokuza/atlassian-server-mcp doctorAny FAIL exits non-zero; WARN never blocks.
Security warning
- TLS certificate and hostname verification is on by default.
--no-tls-verify/ATLASSIAN_TLS_VERIFY=falsedisables it and is only acceptable for local testing against self-signed endpoints — an active network attacker can then impersonate your server. - The exposure tier defaults to
read.--exposure-tier=safe|risky|maxwidens what an LLM agent may invoke; it never replaces Atlassian PAT permissions. - See the Wiki Security Model page for the threat model.
Compatibility
Generation/test baseline (not a startup allowlist — other versions start with a warning and actual compatibility is determined by the target instance):
| Product | Baseline versions | Pinned E2E image |
|---|---|---|
| Jira | DC 10.3 / 11.3 | atlassian/jira-software:11.3.5 |
| Confluence | DC 9.2 / 10.2 | atlassian/confluence:10.2.11 |
| Bitbucket | DC 9.4 / 10.2 / 10.4 | atlassian/bitbucket:10.4.1 |
Node.js >=22 is required. Jira Service Management, Assets, and Marketplace
app APIs are outside v1.
Documentation
User documentation lives in the project Wiki (content source under
docs/wiki/;
the Wiki itself is enabled by the maintainer on deployment):
Project/developer documentation:
docs/en/architecture.md— runtime data flow and sources of truthdocs/en/exposure-policy.md— operation-level exposure policy, generated from official specsdocs/en/tool-contracts.md— metadata-first writes, body templates, error contractdocs/en/test-strategy.md— CI vs. real-product E2E evidence boundaries, plus the local Data Center E2E walkthroughdocs/en/release-process.md— maintainer release memo
Chinese counterparts of every page live under docs/zh-CN/ and
docs/wiki/zh-CN/; the English version is authoritative.
Contributing and support
Development setup and the generated-files discipline are documented in
AGENTS.md
and the developer documentation above. Report bugs and request features
through GitHub issues.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
Neon Database
MCP server for interacting with Neon Management API and databases
E2B
Using MCP to run code via e2b.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.