gouqi

The gouqi MCP server provides an AI-friendly interface for interacting with JIRA without needing to know JQL or the Jira internals. It translates natural-language style, semantic parameters into Jira operations such as searching issues, retrieving issue details, and listing issues assigned to users. Tools exposed by the server include search_issues, get_issue_details, get_user_issues, get_server_status, test_connection, and clear_cache. You can query for issues using semantic predicates (e.g., issue_types, status, project_key, assigned_to) and the server will handle mapping to Jira concepts, applying authentication, and caching results for performance. The integration is designed to work with MCP clients like Claude Desktop and Continue.dev, where you configure the server under the MCP servers section and pass essential environment variables for Jira authentication and connection settings.

Once the server is running, you can test connectivity and tool availability via the provided test_connection and tools list/call endpoints. For example, test_connection will validate Jira access and authentication, while tools/list can reveal available tools and their capabilities. Semantic parameter mappings allow you to specify high-level intents (e.g., find overdue bugs assigned to me) and the server will translate them into Jira queries through its semantic_mapping layer. Ensure your environment variables reflect your Jira deployment (Cloud with PAT tokens, or Server with username/password or API tokens) and consider tuning TTL, max results, and timeouts to suit your workflow.

Prerequisites:

• Rust installed (Rust 1.75.0 or later)
• Access to a Jira instance (Cloud or Server)
• Jira authentication credentials

1. Install Rust and set up the environment

• Install Rust from https://www.rust-lang.org/tools/install
• Ensure rustc and cargo are accessible from your shell

2. Clone the repository and build

git clone https://github.com/yourusername/gouqi-mcp-server.git
cd gouqi-mcp-server
cargo build --release

3. Configure environment variables Create a .env or export variables to point the MCP server at your Jira instance. Example:

export JIRA_URL="https://your-company.atlassian.net"
export JIRA_AUTH_TYPE="pat"  # or "basic", "bearer", "anonymous"
export JIRA_TOKEN="your_token"
export JIRA_CACHE_TTL="300"
export JIRA_MAX_RESULTS="50"
export JIRA_REQUEST_TIMEOUT="30"

4. Run the server

./target/release/jira-mcp-server

5. Optional: test with MCP Inspector or your MCP client in the integration section of the README to verify tools are available.

Notes:

• If you need to run in a different environment, build a release binary and point the mcp config to the binary path.
• For Basic auth, supply JIRA_USERNAME and JIRA_PASSWORD in addition to JIRA_URL and JIRA_AUTH_TYPE.

Tips and caveats:

• Security: never commit credentials; use environment variables or a secret manager.
• If you encounter authentication issues, verify JIRA_TOKEN permissions and the selected JIRA_AUTH_TYPE matches your Jira deployment (PAT for Cloud, username/password or API token for Server).
• Cache TTL (JIRA_CACHE_TTL) can greatly affect latency and API load; tune based on how fresh you need data.
• Ensure network access to the Jira instance from the host running the MCP server; check firewalls and VPNs.
• The server uses semantic parameter mappings; keep your AI prompts close to the supported fields like issue_types, status, project_key, assigned_to, and due_date_filter for predictable results.