Agentic Rust MCP

Production-grade Rust MCP (Model Context Protocol) server for agentic AI pipelines.

Integrates Render, Vercel, Buffer, and Firestore into a unified interface for Claude Code and other AI agents.

✨ All 4 Stages Complete

Stage 1: Foundation ✅

• Async stdio transport
• Message loop architecture
• tokio runtime initialization

Stage 2: Resources & Prompts ✅

• 3 Read-Only Resources (system status, content schedules, activity logs)
• 3 Reusable Prompts (deployment analyzer, content scheduler, activity log analyzer)

Stage 3: Tools ✅

• agency_pulse — Check Render & Vercel deployment status (4 services)
• content_check — Query Buffer content schedules
• data_vault — Access Firestore leads & logs

Stage 4: Roof & Finishings ✅

• Streaming — Real-time updates for long-running tasks
• OAuth 2.1 — Secure token-based authentication
• Professional Logging — JSON structured logs for production
• Tests — 4/4 unit tests passing
• .env Support — Safe credential management

Quick Start

# Build production binary
cargo build --release

# Copy and configure environment
cp .env.example .env
# Edit .env with your OAuth token and API keys

# Run server
./target/release/agentic-rust-mcp

Usage Examples

Request Resources

{"type": "resource", "name": "system-status"}

Request Prompts

{"type": "prompt", "name": "deployment-analyzer"}

Invoke Tools

{"type": "tool", "name": "agency_pulse", "args": {}}
{"type": "tool", "name": "content_check", "args": {}}
{"type": "tool", "name": "data_vault", "args": {}}

Stream Long-Running Task

{"type": "streaming_task", "task": "process_deployment"}

Returns stream markers:

• stream_start — Task initiated
• stream_update — Progress update (3 steps)
• stream_end — Task completed

Security

OAuth 2.1 Token Management

• Load from .env (never hardcoded)
• Verify token validity before requests
• Support token refresh flow
• Bearer token authentication

Credential Protection

• dotenv for safe .env loading
• .env in .gitignore
• .env.example for documentation
• Environment variable support

Performance

Logging

JSON Structured Format

{
  "timestamp": "2026-05-05T00:00:00Z",
  "level": "info",
  "target": "agentic_rust_mcp",
  "message": "Tool executed successfully",
  "tool": "agency_pulse",
  "duration_ms": 125
}

Enable/Disable

RUST_LOG=info ./target/release/agentic-rust-mcp
RUST_LOG=debug ./target/release/agentic-rust-mcp

Architecture

┌─────────────────────────────────────────┐
│  Claude Code / AI Agent                 │
└────────────┬────────────────────────────┘
             │ stdio (JSON)
             ↓
┌─────────────────────────────────────────┐
│  Stage 1: Foundation                    │
│  - Async message loop                   │
│  - Type-safe request parsing            │
└─────────────┬───────────────────────────┘
              │
    ┌─────────┼─────────┐
    ↓         ↓         ↓
┌──────┐  ┌──────┐  ┌──────┐
│ S2   │  │ S3   │  │ S4   │
│─────┐│  │─────┐│  │─────┐│
│Res  ││  │Tool ││  │OAuth││
│Prom ││  │─────││  │Log  ││
│─────││  │agcy ││  │Strm ││
│─────│┘  │cont ││  │─────│┘
│─────│   │data ││  │─────│
└─────┘   └─────┘│  └─────┘
                 │
    ┌────────────┼────────────┐
    ↓            ↓            ↓
[ Render ]  [ Vercel ]   [ Buffer + Firestore ]

Resume Impact

• ✅ Agentic AI Standard — Implements 2026 MCP protocol
• ✅ Production-Ready — Type-safe, async, tested, secured
• ✅ Multi-Stage Build — Shows modular architecture discipline
• ✅ Tool Integration — DevOps + Content + Data orchestration
• ✅ Security First — OAuth 2.1, env vars, no secrets in code
• ✅ Observable — Structured JSON logging for monitoring

Tech Stack

License

MIT

Author

Chris Brown
Albatross AI
Portfolio