spike-mcp
Connects Claude to Jira and Confluence for researching, designing spike docs with Mermaid diagrams, writing Confluence pages, and creating Jira tickets.
README
spike-mcp
MCP server that connects Claude to Jira and Confluence. Engineers can research spikes, generate technical solutions with Mermaid architecture diagrams, write Confluence docs, and create Jira epic + story breakdowns through natural conversation — no Anthropic API key required.
Features
- Research — search Confluence and Jira before generating anything
- Design — produce Mermaid architecture/flow diagrams as part of every spike doc
- Write — create or update Confluence pages (markdown + Mermaid → Confluence storage format)
- Ticket — create Jira Epics, Stories with acceptance criteria, and Tasks; supports both next-gen and classic projects
- Zero LLM cost — Claude itself provides all intelligence; no separate AI API key needed
Installation
pip install spike-mcp
# or
uv add spike-mcp
Setup
Step 1 — Create ~/.spike.toml
Create ~/.spike.toml in your home directory with your Atlassian details:
# macOS / Linux
touch ~/.spike.toml
[atlassian]
base_url = "https://yourorg.atlassian.net"
email = "you@yourorg.com"
[confluence]
# If your Confluence is on a different domain than Jira, set this:
# base_url = "https://yourorg-docs.atlassian.net"
space_key = "ENG"
parent_page_id = "123456"
[jira]
project_key = "PLAT"
epic_issue_type = "Epic"
story_issue_type = "Story"
task_issue_type = "Task"
default_label = "spike"
story_points_field = "customfield_10016"
epic_link_field = "customfield_10014"
[tickets]
story_point_scale = [1, 2, 3, 5, 8, 13]
Tip: Config is also discovered automatically in your project root or any parent directory up to the git root — useful when running spike-mcp from a specific repo. The search order is: current directory → git root walk-up →
~/.spike.toml.
Org-specific required fields
Some Jira projects enforce mandatory custom fields (e.g. Account, Tier, Work type). Add them under [jira.required_fields] and they will be merged into every ticket created:
[jira.required_fields]
customfield_11139 = 15 # plain integer (check your field type)
customfield_11518 = { id = "12203" } # single select
customfield_11664 = [{ id = "13089" }] # multi-select (array)
To find the correct field IDs and allowed values for your project, call the Jira create-meta API:
GET /rest/api/3/issue/createmeta?projectKeys=PROJ&issuetypeNames=Epic&expand=projects.issuetypes.fields
Step 2 — Set your API token
Generate an Atlassian API token at https://id.atlassian.com/manage-profile/security/api-tokens and export it:
export ATLASSIAN_API_TOKEN="your-token-here"
Never put the token in .spike.toml — it is read exclusively from the environment.
Connect to Claude
spike-mcp works with both Claude Desktop (GUI app) and Claude Code (CLI). Follow the guide for whichever you use — or both.
Option A — Claude Desktop
Step 1 — Find the config file
| Platform | Path |
|---|---|
| macOS | ~/Library/Application Support/Claude/claude_desktop_config.json |
| Windows | %APPDATA%\Claude\claude_desktop_config.json |
Step 2 — Add spike-mcp to mcpServers
{
"mcpServers": {
"spike-mcp": {
"command": "spike-mcp",
"env": {
"ATLASSIAN_API_TOKEN": "your-token-here"
}
}
}
}
macOS/Homebrew tip: If
spike-mcpis not on your PATH, use the full binary path:"command": "/opt/homebrew/bin/spike-mcp"Run
which spike-mcpin your terminal to find the path.
Step 3 — Restart Claude Desktop
Quit and reopen the app. A hammer icon (🔨) in the toolbar confirms the tools are active.
Option B — Claude Code (CLI)
Step 1 — Add spike-mcp globally
Run this once so the server is available in every project:
claude mcp add spike-mcp spike-mcp -s user -e ATLASSIAN_API_TOKEN="your-token-here"
macOS/Homebrew tip: Use the full path if
spike-mcpis not on your PATH:claude mcp add spike-mcp /opt/homebrew/bin/spike-mcp -s user -e ATLASSIAN_API_TOKEN="your-token-here"
Step 2 — Verify the connection
claude mcp list
You should see:
spike-mcp: /opt/homebrew/bin/spike-mcp - ✓ Connected
Step 3 — Start a new Claude Code session
MCP servers are loaded at session start. Open a fresh session in any directory and spike-mcp tools will be available automatically.
Example conversations
Start a spike session by invoking the workflow prompt at the top of your conversation:
/mcp__spike-mcp__spike_workflow
Then talk naturally:
- "I need to spike on replacing our job queue with Temporal. Research what we have, design a solution with a diagram, write a spike doc in Confluence under the Platform space, then create an epic and stories in the PLAT project."
- "Search Confluence for our auth service architecture and summarise what you find."
- "Break down the spike doc you just wrote into Jira tickets — 1 epic, stories with acceptance criteria, fibonacci points."
- "What Jira tickets are open in PLAT related to observability?"
Tools
| Tool | Inputs | Description |
|---|---|---|
search_confluence |
query, space_key?, limit? |
CQL full-text search across Confluence |
get_confluence_page |
page_id |
Fetch full page content by ID |
search_jira |
query, project_key?, limit? |
JQL full-text search across Jira |
write_spike_doc |
title, body_markdown, space_key?, parent_page_id? |
Create a Confluence page (markdown + Mermaid → storage format) |
create_epic |
summary, description, project_key?, label? |
Create a Jira Epic |
create_story |
epic_key, summary, description, acceptance_criteria, story_points?, project_key?, label? |
Create a Jira Story linked to an Epic |
create_task |
epic_key, summary, description, project_key?, label? |
Create a Jira Task linked to an Epic |
get_project_config |
— | Show current config targets (API token redacted) |
Configuration reference
All fields in .spike.toml:
| Field | Required | Description |
|---|---|---|
atlassian.base_url |
Yes | Your Atlassian Cloud domain, e.g. https://myorg.atlassian.net |
atlassian.email |
Yes | Your Atlassian account email |
confluence.base_url |
No | Override if Confluence is on a different domain than Jira |
confluence.space_key |
No | Confluence space key for new spike docs, e.g. ENG |
confluence.parent_page_id |
No | Page ID to nest new docs under |
jira.project_key |
No | Jira project key, e.g. PLAT |
jira.epic_issue_type |
No | Issue type name for epics (default: Epic) |
jira.story_issue_type |
No | Issue type name for stories (default: Story) |
jira.task_issue_type |
No | Issue type name for tasks (default: Task) |
jira.default_label |
No | Label applied to all created tickets (default: spike) |
jira.story_points_field |
No | Custom field ID for story points; varies per instance |
jira.epic_link_field |
No | Custom field ID for epic link; classic projects only |
jira.required_fields |
No | Org-specific mandatory fields merged into every create call |
tickets.story_point_scale |
No | Fibonacci scale used when prompting for estimates |
Contributing
git clone https://github.com/de-cryptor/spike-mcp
cd spike-mcp
python -m venv .venv && source .venv/bin/activate
pip install -e ".[dev]"
pytest
License
MIT — see LICENSE.
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.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
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.