atlassian-mcp-server

Provides a Spring Boot MCP (SSE-based) server that performs Atlassian OAuth2 (Jira + Confluence Cloud), stores encrypted tokens in a database, refreshes tokens automatically, and exposes MCP “tools” for Jira issue operations and Confluence page/space operations.

Evaluated Apr 04, 2026 (16d ago)
Repo ↗ Infrastructure mcp spring-boot oauth2 atlassian jira confluence llm-tools sse postgres
⚙ Agent Friendliness
54
/ 100
Can an agent use this?
🔒 Security
67
/ 100
Is it safe for agents?
⚡ Reliability
18
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
80
Documentation
65
Error Messages
0
Auth Simplicity
45
Rate Limits
5

🔒 Security

TLS Enforcement
80
Auth Strength
70
Scope Granularity
60
Dep. Hygiene
50
Secret Handling
70

Security is partially addressed via encrypted token storage (Jasypt) and storing refresh/access tokens in a database. However, the README does not describe TLS enforcement requirements for the app endpoint, detailed token lifecycle/expiry semantics for the connection token, token revocation behavior, CSRF protections for OAuth callback endpoints, database encryption-at-rest controls, or dependency/CVE hygiene. It also instructs enabling logs for raw HTTP response debugging, which can risk leaking sensitive details if log redaction is not implemented.

⚡ Reliability

Uptime/SLA
0
Version Stability
30
Breaking Changes
0
Error Recovery
40
AF Security Reliability

Best When

You can run a local or private Spring Boot service, complete Atlassian OAuth consent for the target user, and then let an agent call the exposed MCP tools using the connection token.

Avoid When

You need turnkey hosted SaaS reliability/SLA, or you can’t securely manage encryption passwords, database credentials, and per-user access tokens.

Use Cases

  • Jira issue retrieval/creation/update via an LLM agent
  • Confluence page search and content retrieval/summarization via an LLM agent
  • Creating Confluence pages from agent-generated content
  • Building agent workflows that act on a user’s Atlassian account after OAuth consent

Not For

  • Multi-tenant environments without strong isolation controls for per-principal tokens
  • Highly compliance-sensitive deployments that require documented security posture and auditability beyond the README
  • Public unauthenticated exposure of the MCP endpoint without additional network controls

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
Yes
SDK
No
Webhooks
No

Authentication

Methods: OAuth2 authorization_code (Atlassian) Bearer Authorization header using an Atlassian connection token (principalName/UUID) for MCP tool calls
OAuth: Yes Scopes: Yes

Flow described includes Atlassian OAuth2 token exchange with offline_access for refresh tokens; tool access is mediated by a short-lived connection token shown post-consent.

Pricing

Free tier: No
Requires CC: No

No pricing information provided; appears self-hosted (infrastructure cost = your compute + DB + Atlassian API usage under your Atlassian app limits).

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • OAuth scopes must be correctly configured for Jira/Confluence; 401s may indicate missing/incorrect scopes
  • The agent must use the connection token (principalName/UUID) produced after OAuth success; expired/incorrect tokens will fail tool calls
  • The README references enabling logs for debugging raw Confluence HTTP responses; agents may need log visibility for troubleshooting
  • MCP transport is SSE; ensure the agent/client supports SSE and that endpoint access (including headers like Authorization) is configured correctly

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for atlassian-mcp-server.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-04-04.

8642
Packages Evaluated
17761
Need Evaluation
586
Need Re-evaluation
Community Powered