OrangePro — Claude Code MCP server: test analysis and setup
OrangePro finds functions without proven test protection, then checks tests by changing code in an isolated copy. Use it to see which testing gaps deserve attention first.
claude mcp add --env OPENAI_API_KEY=<OPENAI_API_KEY> --env ANTHROPIC_API_KEY=<ANTHROPIC_API_KEY> --env OLLAMA_BASE_URL=<OLLAMA_BASE_URL> --transport stdio orangepro -- npx -y @orangepro/mcp-serverWhat OrangePro does
The server maps public repository functions to tests that call them, then orders uncovered areas by potential impact. To verify a test, it replaces a function body in an isolated copy and runs that test. If the test still passes, the function’s behavior is not proven to be protected. Analysis and ranking run locally and need no model key.
OrangePro writes two HTML reports and tests.csv to .orangepro/. The reports show a prioritized worklist and how tests reach code; the CSV contains draft test steps, data, and expected outcomes. Draft generation for top-ranked gaps uses your own model key. Install the repository’s dependencies first, or the project analysis cannot run.
What it helps with
- find public functions without proven test protection
- tell a function call apart from a behavioral check
- check whether a test fails after its function is changed
- rank test gaps by potential impact
- export draft test cases as CSV
Example prompts for the agent
- Analyze this repository’s tests and list functions without proven protection.
- Check the test for
deleteUser: change the function in an isolated copy and run the test. - Draft tests for the highest-priority gaps and write the reports to
.orangepro/.
How to install OrangePro
Platform CLI
claude mcp add --env OPENAI_API_KEY=<OPENAI_API_KEY> --env ANTHROPIC_API_KEY=<ANTHROPIC_API_KEY> --env OLLAMA_BASE_URL=<OLLAMA_BASE_URL> --transport stdio orangepro -- npx -y @orangepro/mcp-serverReplace <…> with your own values
Platform CLI
codex mcp add orangepro --env OPENAI_API_KEY=<OPENAI_API_KEY> --env ANTHROPIC_API_KEY=<ANTHROPIC_API_KEY> --env OLLAMA_BASE_URL=<OLLAMA_BASE_URL> -- npx -y @orangepro/mcp-serverReplace <…> with your own values
Config file
{
"mcpServers": {
"orangepro": {
"command": "npx",
"args": [
"-y",
"@orangepro/mcp-server"
],
"env": {
"OPENAI_API_KEY": "<OPENAI_API_KEY>",
"ANTHROPIC_API_KEY": "<ANTHROPIC_API_KEY>",
"OLLAMA_BASE_URL": "<OLLAMA_BASE_URL>"
}
}
}
}Replace <…> with your own values
Config file
{
"servers": {
"orangepro": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@orangepro/mcp-server"
],
"env": {
"OPENAI_API_KEY": "<OPENAI_API_KEY>",
"ANTHROPIC_API_KEY": "<ANTHROPIC_API_KEY>",
"OLLAMA_BASE_URL": "<OLLAMA_BASE_URL>"
}
}
}
}Replace <…> with your own values
Platform CLI
gemini mcp add -e OPENAI_API_KEY=<OPENAI_API_KEY> -e ANTHROPIC_API_KEY=<ANTHROPIC_API_KEY> -e OLLAMA_BASE_URL=<OLLAMA_BASE_URL> orangepro npx -y @orangepro/mcp-serverReplace <…> with your own values
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about OrangePro
How do I install OrangePro in Claude Code?
Run claude mcp add mcp -e OPENAI_API_KEY=<OPENAI_API_KEY> -e ANTHROPIC_API_KEY=<ANTHROPIC_API_KEY> -- npx -y @orangepro/mcp-server. Replace the placeholders with key values if you plan to generate tests.
What does OrangePro do?
It maps repository functions to tests that call them, ranks gaps, and checks test protection by changing a function in an isolated copy and rerunning the test.
Does project analysis require an API key?
Analysis, ranking, and test proof work without a model key. Draft test generation and suggestions use your own key.
