lmsnex-mcp-server

MCP bridge between NEXA agents and LMSNEX, supporting three agent domains:

Architecture

┌─────────────┐   MCP    ┌──────────────────┐   REST   ┌──────────┐
│  NEXA Agent │ ◄──────► │  lmsnex-server   │ ◄──────► │  LMSNEX  │
│  (client)   │ stdio/SSE│  (Python)        │  x-api-key│  API     │
└─────────────┘          └──────────────────┘          └──────────┘

See ARCHITECTURE.md for details and AGENTS.md for per-agent tool scoping.

Quick start

1. Install

cd lmsnex-mcp-server
python -m venv venv
source venv/bin/activate   # Windows: venv\Scripts\activate
pip install -r requirements.txt
cp app/config.example.py app/config.py

2. Configure

Set environment variables:

export LMSNEX_API_BASE_URL="https://your-lmsnex-host/api"
export LMSNEX_API_KEY="your-api-key-here"

3. Run (stdio)

python -m app.mcp_server
# or after pip install:
mcp-lmsnex

4. Run (HTTP — Cloud Run / local)

./run-http-server-local.sh
# Health: http://localhost:8080/health
# Tools:  GET http://localhost:8080/mcp/tools

NEXA MCP client config (stdio)

{
  "mcpServers": {
    "lmsnex": {
      "command": "/path/to/venv/bin/python",
      "args": ["-m", "app.mcp_server"],
      "cwd": "/path/to/lmsnex-mcp-server",
      "env": {
        "LMSNEX_API_BASE_URL": "https://your-lmsnex-host/api",
        "LMSNEX_API_KEY": "your-api-key-here"
      }
    }
  }
}

MCP tools

Creator

• upload_source_material, ingest_source_material, create_course_draft
• generate_course_modules, save_course_modules
• save_course_objectives, save_course_assessments
• submit_course_for_review, get_course_draft_status
• generate_course_from_article, generate_course_from_powerpoint (text-only PPTX → Moodle)
• generate_quiz_from_content (course text → MCQ / T-F / short answer; easy|medium|hard)
• generate_speaker_notes, split_presentation_into_chapters / split_master_deck

Manager

• list_users, list_cohorts, enroll_users, assign_learning_path
• get_enrollment_report, get_completion_report
• issue_certification, get_certification_status

Learner

• search_course_content (returns moduleId, title, excerpt citations)
• get_module_content, get_learner_progress
• get_recommended_next_steps, record_knowledge_gap

MCP resources

• lmsnex://courses/catalog
• lmsnex://learning-paths/all

Docker

docker build -t lmsnex-mcp-server .
docker run -p 8080:8080 \
  -e LMSNEX_API_BASE_URL="https://your-lmsnex-host/api" \
  -e LMSNEX_API_KEY="your-api-key-here" \
  lmsnex-mcp-server

Cloud Run

./deploy-cloudrun-http.sh

Tests

python test_api.py
python test_mcp_client.py
python test_setup.py
python test_package.py

Project layout

lmsnex-mcp-server/
├── app/
│   ├── mcp_server.py          # Stdio MCP server
│   ├── http_mcp_server.py     # HTTP/SSE wrapper
│   ├── config.example.py
│   └── loaders/
│       └── lmsnex_api_loader.py
├── pyproject.toml
├── requirements.txt
├── Dockerfile
├── AGENTS.md
└── ARCHITECTURE.md

License

MIT