Containerized PMC Execution Engine

This repository illustrates how to bridge the Model Command Protocol (PMC) with the capabilities of OpenAI's service, allowing the LLM to leverage utilities exposed by a PMC engine operating within a Docker container.

Prerequisites

• A functional Docker installation on your host system
• Git version control utility (for repository cloning)

Component Layout

• server.py: Contains the PMC engine logic and utility definitions
• client.py: Script for establishing connection to the engine and invoking the agent
• Dockerfile: Blueprint for constructing the necessary container image
• requirements.txt: List of Python packages essential for the project

Operational Sequence Detail

1. Inquiry Submission: The user presents a prompt to the system (e.g., "What is the established protocol for employee leave?")
2. OpenAI Processing: OpenAI ingests the prompt alongside the exposed utility descriptions from the PMC gateway.
3. Capability Selection: The model determines the most appropriate utility (or utilities) to address the prompt.
4. PMC Invocation Relay: The client component receives the designated utility call instruction from OpenAI and relays it to the PMC engine.
5. Engine Execution: The PMC engine executes the designated function (e.g., querying an internal repository).
6. Result Forwarding: The output generated by the utility travels back via the client interface to the OpenAI service.
7. Final Synthesis: OpenAI formulates the conclusive answer, incorporating the retrieved utility data.

Deployment Via Docker

Phase 1: Image Construction

docker build -t mog-service .

Phase 2: Container Launch

docker run -p 8050:8050 mog-service

This command initializes the PMC engine within a sandboxed container, making its interface available externally on TCP port 8050.

Client Execution

Once the service is confirmed operational, execute the client script in a separate terminal session:

python client.py

The client will establish a connection, enumerate available functions, and initiate the agent process to resolve the initial query.

Diagnostic Guidance

If connectivity challenges arise:

1. Service Status Verification: Confirm the container is active using docker ps.

2. Port Mapping Validation: Review the output of docker ps or the docker run command summary to confirm correct port forwarding.

3. Engine Log Inspection: Examine container outputs via docker logs <container_identifier> for runtime errors.

4. Interface Binding: The engine is bound to 0.0.0.0 for external access. If issues persist, examine local firewall configurations.

5. Network Topology: If the Docker host is remote, ensure that port 8050 is reachable from the client machine's network segment.

Remarks

• The communication layer utilizes Server-Sent Events (SSE) streaming, listening specifically on port 8050.
• The client targets the endpoint at http://localhost:8050/sse.
• Confirm the engine is fully initialized before attempting to run the client application.