Bring Web Server
to Pydantic AI
Learn how to connect Caddy Server to Pydantic AI and start using 13 AI agent tools in minutes. Fully managed, enterprise secure, and ready to use without writing a single line of code.
GDPR Free for SubscribersCompatible with every major AI agent and IDE
Gemini
What is the Caddy Server MCP Server?
Connect your Caddy Server instance to any AI agent and automate your web infrastructure management through natural conversation.
What you can do
- Configuration Management — Load, get, append, or replace server configurations using JSON or Caddyfile formats.
- Caddyfile Adaptation — Instantly convert Caddyfile text into native Caddy JSON without applying the changes.
- Upstream Monitoring — Check the real-time status and health of your proxy upstreams and backends.
- PKI & Certificates — Inspect internal CA information and retrieve certificate chains for your managed domains.
- Metrics & Observability — Access Prometheus-style metrics to monitor server performance and request traffic.
- Granular Control — Delete specific configuration paths or gracefully stop the server process.
How it works
- Subscribe to this server
- Enter your Caddy Admin API URL (e.g., http://localhost:2019)
- Start managing your reverse proxies and web servers from Claude, Cursor, or any MCP client
Who is this for?
- DevOps Engineers — automate infrastructure updates and monitor backend health without leaving the terminal or chat.
- Web Developers — quickly test and adapt Caddyfile configurations during local development.
- SREs — retrieve live metrics and PKI status to ensure site reliability and security compliance.
Built-in capabilities (13)
Adapts a configuration (e.g., Caddyfile) to JSON without running it
and target is array, expands payload array and appends elements. Sets or replaces an object; appends to an array in Caddy config
Deletes the value at the named path in Caddy config
Leave empty for full config. Exports the configuration at the specified path as JSON
Access a configuration object directly via its @id field
Exposes metrics in Prometheus exposition format
Returns information about a particular PKI app CA
Returns the certificate chain for a particular CA
Returns the current status of configured proxy upstreams
Creates a new object or inserts into an array at a specific index
Use application/json for native JSON, or text/caddyfile for Caddyfile. Sets or replaces the active Caddy configuration
Strictly replaces an existing object or array element in Caddy config
Gracefully shuts down the Caddy server and exits the process
Why Pydantic AI?
Pydantic AI validates every Caddy Server tool response against typed schemas, catching data inconsistencies at build time. Connect 13 tools through Vinkius and switch between OpenAI, Anthropic, or Gemini without changing your integration code. full type safety, structured output guarantees, and dependency injection for testable agents.
- —
Full type safety: every MCP tool response is validated against Pydantic models, catching data inconsistencies before they reach your application
- —
Model-agnostic architecture. switch between OpenAI, Anthropic, or Gemini without changing your Caddy Server integration code
- —
Structured output guarantee: Pydantic AI ensures tool results conform to defined schemas, eliminating runtime type errors
- —
Dependency injection system cleanly separates your Caddy Server connection logic from agent behavior for testable, maintainable code
Caddy Server in Pydantic AI
Caddy Server and 4,000+ other MCP servers. One platform. One governance layer.
Teams that connect Caddy Server to Pydantic AI through Vinkius don't need to source, host, or maintain individual MCP servers. Every tool call runs inside a hardened runtime with credential isolation, DLP, and a signed audit chain.
Raw MCP | Vinkius | |
|---|---|---|
| Server catalog | Find and host yourself | 4,000+ managed |
| Infrastructure | Self-hosted | Sandboxed V8 isolates |
| Credential handling | Plaintext in config | Vault + runtime injection |
| Data loss prevention | None | Configurable DLP policies |
| Kill switch | None | Global instant shutdown |
| Financial circuit breakers | None | Per-server limits + alerts |
| Audit trail | None | Ed25519 signed logs |
| SIEM log streaming | None | Splunk, Datadog, Webhook |
| Honeytokens | None | Canary alerts on leak |
| Custom domains | Not applicable | DNS challenge verified |
| GDPR compliance | Manual effort | Automated purge + export |
Why teams choose Vinkius for Caddy Server in Pydantic AI
The Caddy Server MCP Server runs on Vinkius-managed infrastructure inside AWS — a purpose-built runtime with per-request V8 isolates, Ed25519 signed audit chains, and sub-40ms cold starts. All 13 tools execute in hardened sandboxes optimized for native MCP execution.
Your AI agents in Pydantic AI only access the data you authorize, with DLP that blocks sensitive information from ever reaching the model, kill switch for instant shutdown, and up to 60% token savings. Enterprise-grade infrastructure, zero maintenance.
* Every MCP server runs on Vinkius-managed infrastructure inside AWS - a purpose-built runtime with per-request V8 isolates, Ed25519 signed audit chains, and sub-40ms cold starts optimized for native MCP execution. See our infrastructure
How Vinkius secures
Caddy Server for Pydantic AI
Every tool call from Pydantic AI to the Caddy Server MCP Server is protected by DLP redaction, cryptographic audit chains, V8 sandbox isolation, kill switch, and financial circuit breakers.
Frequently asked questions
Can I use a Caddyfile instead of JSON to update my configuration?
Yes! Use the load_config tool and set the content_type to text/caddyfile. You can also use adapt_config to preview the JSON conversion before applying it.
How can I monitor the health of my load-balanced backends?
Use the get_upstreams tool. It returns the current status and health metrics of all configured proxy upstreams in your Caddy instance.
Is it possible to remove a specific site or route without resetting the whole server?
Absolutely. Use delete_config with the specific path (e.g., apps/http/servers/srv0/routes/1) to remove only that element from the active configuration.
How does Pydantic AI discover MCP tools?
Create an MCPServerHTTP instance with the server URL. Pydantic AI connects, discovers all tools, and generates typed Python interfaces automatically.
Does Pydantic AI validate MCP tool responses?
Yes. When you define result types as Pydantic models, every tool response is validated against the schema. Invalid data raises a clear error instead of silently corrupting your pipeline.
Can I switch LLM providers without changing MCP code?
Absolutely. Pydantic AI abstracts the model layer. your Caddy Server MCP integration works identically with OpenAI, Anthropic, Google, or any supported provider.
MCPServerHTTP not found
Update: pip install --upgrade pydantic-ai
Explore More MCP Servers
View all →
Federal Register API
4 toolsAccess US public documents — audit rules and agencies via AI.
arXiv
2 toolsSearch 2.5M+ scientific preprints across physics, computer science, mathematics, biology, economics, and more — with full abstracts, author lists, and direct PDF download links.
ANEEL Dados Abertos
4 toolsAccess the Brazilian National Electric Energy Agency (ANEEL) Open Data Portal to query electricity sector datasets, resources, and records.
Maxio (SaaS Billing & FinOps)
10 toolsManage SaaS billing via Maxio — audit customer contracts, track ARR/MRR metrics, and create invoices.