bnomei/kirby-mcp

Description and use

**Brief Summary**
bnomei/kirby‑mcp is a CLI‑first “MCP” (multi‑agent coordination platform) server built for Composer‑based Kirby CMS projects. It lets developers inspect blueprints, templates, and plugins, run a live Kirby runtime, and tap into a bundled Kirby knowledge base, turning ad‑hoc prompts and tools into repeatable, orchestrated agent workflows.

**Value**

• **Unified workflow engine** – By exposing Kirby’s internal structures (blueprints, templates, plugins) through a programmable API/CLI, the project lets AI agents query, modify, and validate a Kirby site just like a human developer would.
• **Tool‑use pipelines** – Agents can call the MCP server as a tool, chaining actions (e.g., “generate a new template → validate against blueprint → commit to Git”) without custom scripting for each step.
• **Standardised memory** – The bundled knowledge base provides a persistent, domain‑specific context that agents can reference, reducing prompt drift and improving consistency across runs.

**Practical Adoption Path**

1. **Prototype** – Clone the repo, run composer install and start the MCP server locally (php bin/mcp serve). Use the provided CLI to query a Kirby site’s blueprints and verify responses.
2. **Integrate** – Wrap the MCP API in your existing AI orchestration layer (e.g., LangChain, CrewAI). Define tool specs that call the MCP endpoints for “inspect‑blueprint”, “render‑template”, etc.
3. **Automate** – Build multi‑agent pipelines that combine MCP calls with other tools (code generation, CI/CD, content linting). Store the MCP knowledge base in a vector store for RAG if you need richer natural‑language retrieval.
4. **Deploy** – Containerise the server (Dockerfile is included) and run it as a side‑car in your Kirby production stack, exposing only internal network access to keep the surface area small.

**Production Readiness**

• **Activity & Community** – 56 stars, recent commits (as of 2026‑06‑29), and a modest but active fork base indicate ongoing maintenance.
• **Technical maturity** – The project ships a stable CLI, a well‑documented API, and a PHP codebase that aligns with Kirby’s Composer workflow, making integration straightforward for PHP teams.
• **Risk considerations** – License and security audits are still pending; verify the license (likely MIT) and run a dependency scan before production. No major metadata gaps were found, and the architecture is simple enough to sandbox.

Overall, bnomei/kirby‑mcp is production‑ready for pilot projects, offering a low‑friction way to embed Kirby‑specific knowledge and actions into AI‑driven automation pipelines.

Use in a workflow

bnomei/kirby-mcp helps turn isolated prompts and tools into repeatable agent workflows.

• coordinate multi-agent workflows
• add tool-use pipelines
• standardize agent memory

Limitations and review notes

High for an OSS candidate: recent activity, adoption, and ecosystem signals are strong enough for a serious pilot.