MCP Integrations for AI Coding Agents
ContextsBase connects your coding agent to your product backlog, data schemas, business rules, and test cases over the open Model Context Protocol (MCP). Choose your agent environment to get started.
Cursor
Connect via .cursor/mcp.json. Cursor claims iteration items, ingests JSON data models, and executes backlog tasks without .cursorrules bloat.
// .cursor/mcp.json
{
"mcpServers": {
"contextsbase": {
"command": "npx",
"args": ["-y", "@contextsbase/mcp-server"]
}
}
}Claude Code
Connect with claude mcp add. Run multi-step feature iterations directly from your terminal with persistent specs and Playwright test recording.
$ claude mcp add contextsbase \ -- npx -y @contextsbase/mcp-server \ --env CONTEXTSBASE_TOKEN="cb_tok_..."
Windsurf
Configure via mcp_config.json. Windsurf Cascade pulls Given/When/Then test cases and architecture tokens for deep code synthesis.
// mcp_config.json
{
"mcpServers": {
"contextsbase": {
"command": "npx",
"args": ["-y", "@contextsbase/mcp-server"]
}
}
}GitHub Copilot
Bridge Copilot directly to your product specs. Keep models aligned to the single source of truth across team repositories.
// VS Code settings.json
{
"github.copilot.mcpServers": {
"contextsbase": { ... }
}
}Need a custom or internal MCP setup?
ContextsBase implements standard JSON-RPC 2.0 over stdio and HTTP SSE. Any custom agent or framework that supports the Model Context Protocol can connect with a single API token.