sdlc-assist-mcp
Generates IT cost estimations by calling Vertex AI Gemini directly with project context.
Provides read access to SDLC project artifacts stored in Supabase, including PRDs, architecture docs, data models, API contracts, screen inventories, and tech preferences.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@sdlc-assist-mcp What projects do I have?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
SDLC Assist MCP Server
An MCP (Model Context Protocol) server that gives AI assistants read access to your SDLC Assist project artifacts stored in Supabase.
What This Does
When connected to Claude Desktop or Claude Code, this server lets you have conversations about your SDLC projects:
"What projects do I have?"
"Show me the data model for the DEP Multi-Tenant project"
"What API endpoints handle authentication?"
"List all the screens for the HCP Portal"
"What tech stack did we choose?"
The AI reads your project data directly from Supabase — PRDs, architecture docs, data models, API contracts, screen inventories, and more. It can also generate IT cost estimations by calling Vertex AI Gemini directly with project context.
Related MCP server: Supabase MCP Server
How MCP Works (Quick Primer)
You (in Claude Desktop)
│ "What does the data model look like for DEP Multi-Tenant?"
│
▼
Claude (the AI)
│ Thinks: "I need the data model artifact for that project"
│ Calls: sdlc_get_artifact(project_id="dc744778...", artifact_type="data_model")
│
▼
This MCP Server
│ Queries Supabase for the data_model_content column
│ Returns the full markdown document
│
▼
Claude (the AI)
│ Reads the data model, answers your question
▼
You see the answerMCP is just a protocol — a standardized way for AI to call functions. This server exposes 6 tools that the AI can call when it needs project data.
Available Tools
Tool | What it does |
| Lists all projects with completion status |
| Detailed overview of one project (artifacts, screens, files) |
| Fetches any artifact: PRD, architecture, data model, API contract, sequence diagrams, implementation plan, CLAUDE.md, or corporate guidelines |
| Lists UI screens with metadata, optionally includes HTML prototypes |
| Returns the tech stack choices for a project |
| Generates Traditional vs AI-Assisted IT cost estimates by calling Vertex AI Gemini directly with project context. Requires all upstream artifacts (PRD, architecture, data model, API contract, implementation plan) to be generated first. |
Architecture
┌─────────────────────────────────────┐
│ MCP Client (Claude) │
└──────────────┬──────────────────────┘
│ MCP Protocol
▼
┌─────────────────────────────────────┐
│ sdlc-assist-mcp Server │
│ (FastMCP · streamable-http/stdio) │
├──────────────┬──────────────────────┤
│ Read Tools │ Gemini Tools │
│ (1-5) │ (6) │
└──────┬───────┴──────────┬───────────┘
│ │
▼ ▼
┌──────────────┐ ┌───────────────────┐
│ Supabase │ │ Vertex AI Gemini │
│ PostgREST │ │ (generateContent │
│ (httpx) │ │ via REST API) │
└──────────────┘ └───────────────────┘Prerequisites
Python 3.10+
uv (recommended) or pip
A Supabase project with the SDLC Assist schema
Claude Desktop or Claude Code
(For estimation tool) Google Cloud project with Vertex AI Gemini API enabled
Setup
1. Clone and install
git clone https://github.com/ramseychad1/sdlc-assist-mcp.git
cd sdlc-assist-mcp
# Using uv (recommended)
uv sync
# Or using pip
pip install -e .2. Configure environment
cp .env.example .envEdit .env with your credentials:
# Required — Supabase
SUPABASE_URL=https://mtzcookrjzewywyirhja.supabase.co
SUPABASE_SERVICE_ROLE_KEY=your-service-role-key-here
# Optional — Vertex AI Gemini (only needed for sdlc_generate_estimation)
VERTEXAI_PROJECT_ID=sdlc-assist
VERTEXAI_LOCATION=us-central1Find your Supabase service role key in: Supabase Dashboard → Settings → API → service_role (secret)
3. Test it works
# Quick syntax check
python -c "from sdlc_assist_mcp.server import mcp; print('Server loads OK')"4. Connect to Claude Desktop
Edit your Claude Desktop config file:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
Add this to the mcpServers section:
{
"mcpServers": {
"sdlc-assist": {
"command": "uv",
"args": [
"run",
"--directory", "/ABSOLUTE/PATH/TO/sdlc-assist-mcp",
"sdlc-assist-mcp"
]
}
}
}Or if using pip instead of uv:
{
"mcpServers": {
"sdlc-assist": {
"command": "/ABSOLUTE/PATH/TO/sdlc-assist-mcp/.venv/bin/sdlc-assist-mcp"
}
}
}Restart Claude Desktop. You should see the SDLC Assist tools in the tools menu.
5. Connect to Claude Code (Antigravity IDE)
claude mcp add sdlc-assist -- uv run --directory /ABSOLUTE/PATH/TO/sdlc-assist-mcp sdlc-assist-mcpProject Structure
sdlc-assist-mcp/
├── pyproject.toml # Dependencies + entry point
├── Dockerfile # Cloud Run container image
├── deploy.sh # GCP deployment script
├── .env.example # Environment template
├── .gitignore
├── README.md
├── src/
│ └── sdlc_assist_mcp/
│ ├── __init__.py
│ ├── server.py # MCP server + all 6 tool definitions
│ ├── supabase_client.py # Async Supabase REST client (httpx)
│ ├── vertex_client.py # Async Vertex AI Gemini client (REST API)
│ └── models/
│ ├── __init__.py
│ └── inputs.py # Pydantic input models for tools
└── tests/
└── (coming soon)Deployment
The server supports two transports:
stdio (default) — For local use with Claude Desktop / Claude Code
streamable-http — For remote deployment on Cloud Run
Deploy to Cloud Run
./deploy.shThis builds the container with Cloud Build, stores Supabase credentials in Secret Manager, and deploys to Cloud Run. See deploy.sh for full details.
Environment Variables (Cloud Run)
Variable | Required | Description |
| Yes | Supabase project URL |
| Yes | Supabase service role key (stored in Secret Manager) |
| For estimation tool | GCP project name (defaults to |
| For estimation tool | GCP region (defaults to |
Future Enhancements
Write tools — Update PRDs, add screens, modify artifacts
More Gemini-powered tools — Route additional generative tasks through Vertex AI Gemini
Search across artifacts — Find mentions of a term across all project documents
Project creation — Start new projects from the chat interface
Available Tools
6 toolssdlc_generate_estimationAIdempotent
Generate Traditional vs AI-Assisted cost estimates for a project.
Produces two side-by-side estimates showing hours and costs for each SDLC phase (Requirements, Design, Develop, Test, Deploy, Data Cleansing, Transition to Run, Project Management), then highlights the savings from using SDLC-Assist + agentic development.
Requires all upstream artifacts to be generated first (PRD, architecture, data model, API contract, screens, implementation plan).
Args: project_id: The project UUID.
Returns: str: JSON with traditionalEstimate, aiAssistedEstimate, savings, and assumptions.
Inputs: params (required). Returns: result
sdlc_get_artifactARead-onlyIdempotent
Fetch the full content of a specific artifact from a project.
Retrieves artifacts like the PRD, Architecture Overview, Data Model, API Contract, Sequence Diagrams, Implementation Plan, CLAUDE.md, or Corporate Guidelines.
Args: params (GetArtifactInput): Contains: - project_id (str): UUID of the project - artifact_type (ArtifactType): Which artifact to fetch. One of: 'prd', 'design_system', 'architecture', 'data_model', 'api_contract', 'sequence_diagrams', 'implementation_plan', 'claude_md', 'corporate_guidelines'
Returns: str: The full artifact content (Markdown or JSON depending on type), or an error message if the artifact hasn't been generated yet.
Inputs: params (required). Returns: result
sdlc_get_project_summaryARead-onlyIdempotent
Get a detailed summary of a single SDLC project.
Returns the project name, status, tech stack preferences, which artifacts have been generated (and when), and the number of UI screens. Does NOT return artifact content — use sdlc_get_artifact for that.
Args: params (GetProjectSummaryInput): Contains: - project_id (str): UUID of the project
Returns: str: Markdown-formatted project summary with artifact status.
Inputs: params (required). Returns: result
sdlc_get_screensARead-onlyIdempotent
List all UI screens defined for a project with their metadata.
Returns screen names, types, complexity, epic assignments, and design notes. Optionally includes the full HTML prototype content.
Args: params (GetScreensInput): Contains: - project_id (str): UUID of the project - include_prototypes (bool): If true, include HTML content (default false — prototypes can be very large)
Returns: str: Markdown-formatted screen inventory grouped by epic.
Inputs: params (required). Returns: result
sdlc_get_tech_preferencesARead-onlyIdempotent
Fetch the technology stack preferences for a project.
Returns the user's selected frontend, backend, database, deployment target, authentication method, and API style choices.
Args: params (GetTechPreferencesInput): Contains: - project_id (str): UUID of the project
Returns: str: Markdown-formatted tech stack summary, or a message indicating preferences haven't been set yet.
Inputs: params (required). Returns: result
sdlc_list_projectsARead-onlyIdempotent
List all SDLC Assist projects with their status and artifact completion.
Returns a summary of every project including which artifacts have been generated (PRD, Architecture, Data Model, etc.) and how many UI screens exist. Use this to discover project IDs for other tools.
Args: params (ListProjectsInput): Optional filters: - status_filter (Optional[str]): Filter by status ('DRAFT', 'ACTIVE', 'COMPLETED', 'ARCHIVED')
Returns: str: Markdown-formatted list of projects with completion info.
Inputs: params (required). Returns: result
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 6 tools
Each tool targets a distinct resource/action: listing projects, getting a project summary, fetching a specific artifact, listing screens, getting tech preferences, and generating estimation. There is no overlap or ambiguity in their purposes.
All tools follow the consistent pattern sdlc_<verb>_<noun> with lower_snake_case: sdlc_list_projects, sdlc_get_project_summary, sdlc_get_artifact, sdlc_get_screens, sdlc_get_tech_preferences, sdlc_generate_estimation. This is uniform and predictable.
Six tools is an appropriate, well-scoped set for a read-and-estimate SDLC assist server. Each tool earns its place, and the count is within the ideal 3-15 range, making the server easy to navigate.
The tool surface covers the core read workflows: listing projects, retrieving any artifact, screen inventory, tech preferences, and generating estimates. Minor gaps exist, such as no tool to create or update projects/artifacts, but the stated purpose appears to be querying and estimation, so this is a reasonable limitation.
Maintenance
Related MCP Connectors
Your Supabase account in natural language: run SQL, apply migrations, manage tables, storage, edge f
Safe, read-only Postgres and MySQL access for AI agents. Audit log + column-level controls.
Supabase RLS simulator, app key-leak check, email sender check, DMARC and payment-fraud checks.
Query your team's drift, vulnerability, and upgrade data from any AI assistant. OAuth 2.1, 51 tools.
Related MCP Servers
- AlicenseAqualityDmaintenanceEnables AI assistants to interact with PostgreSQL databases using natural language queries, providing secure read-only access to database schemas and SQL translation capabilities.6ISC
- AlicenseNot gradedqualityDmaintenanceConnects AI assistants to Supabase projects, enabling them to manage tables, query data, deploy Edge Functions, handle migrations, and access project resources through natural language commands.Apache 2.0
- AlicenseNot gradedqualityDmaintenanceConnects Supabase projects to AI assistants, enabling them to manage database tables, execute SQL queries, and deploy Edge Functions through natural language. It provides a comprehensive suite of tools for project administration, including logs access, documentation search, and environment configuration.Apache 2.0
- FlicenseNot gradedqualityCmaintenanceEnables AI agents to safely interact with Supabase through read-only and gated-write tools, with bounded queries, allowlists, and explicit project targeting.-