OSGi Error Interpreter Lambda

This project exposes an OSGi error interpreter two ways:

• A POST /errorInterpreter HTTP endpoint used by the HTML front-end.
• A POST /mcp Model Context Protocol endpoint that lets LLMs (such as Claude) call the interpreter as a tool when they encounter an OSGi error.

MCP server

The /mcp route is a stateless MCP server that speaks the Streamable HTTP transport using single JSON responses (it does not open server-initiated SSE streams). It exposes one tool:

• interpret_osgi_error – input { "errorMessage": string }; returns a plain-text interpretation plus generic and Elastic Path Self-Managed Commerce remediation guidance.

Connect a client

For Claude Code:

claude mcp add --transport http osgi-error-interpreter <api-base-url>/mcp

Other clients accept the same URL as a remote/HTTP MCP server. No local installation is required.

Test the MCP endpoint locally

The samples/mcp-*.json files contain ready-to-use JSON-RPC requests in API Gateway proxy format:

sls invoke local --function mcp --path samples/mcp-initialize.json --verbose
sls invoke local --function mcp --path samples/mcp-tools-list.json --verbose
sls invoke local --function mcp --path samples/mcp-tools-call.json --verbose

Build

mvn clean install

Test Lambda locally

To run the Lambda locally without deploying to AWS, use these steps:

1. Install the Serverless Framework’s Open Source CLI.
2. Invoke the Lambda functions locally as follows:

   sls invoke local --function errorInterpreter --path samples/sample1.json --verbose

3. To run the commands with remote debugging enabled, run the following before the sls invoke local command:

   export JAVA_TOOL_OPTIONS="-Xdebug -Xrunjdwp:transport=dt_socket,server=y,suspend=y,address=8000 -Xnoagent -Djava.compiler=NONE"

Deploy to AWS

1. Install the AWS Command Line Interface.
2. Run aws configure and enter your AWS account details.
3. Deploy the Lambda using sls deploy.