Ae ltd mcp foundry

Skill aegntic/cldcde/skills/ae-ltd-mcp-foundry

Use when scaffolding or extending MCP servers, adding tools/resources/prompts, or hardening MCP tool schemas and handlers. AE.LTD workflow tuned for production-ready MCP delivery.From its SKILL.md

Install

npx -y skills add aegntic/cldcde --skill ae-ltd-mcp-foundry

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

11 stars11 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

1.8 KB, as published. Nobody here has run it

AE.LTD MCP Foundry

What This Skill Does

Scaffolds and hardens MCP (Model Context Protocol) servers with strict contracts and validation. Niche gem source: model-context-protocol plugin from awesome-claude-code-toolkit.

Scaffold new MCP servers with proper structure
Add tools, resources, and prompts to existing servers
Validate input/output with Zod (TS) or Pydantic (Python)
Test happy-path and failure-path scenarios

Quick Start

User: "Create an MCP server for file operations"
→ Confirm context → Lock contracts → Scaffold → Validate → Test → Document

Workflow

Confirm server context. Path, language, transport (stdio or sse), and whether this is new or existing.
Lock contract first. Define tool names, descriptions, and input/output schema before writing handlers.
Scaffold or extend. Create server structure (if new), then add tool/resource registration and handler wiring.
Validate hard. Use Zod (TS) or Pydantic (Python). Reject invalid input early with useful error text.
Test one happy-path and one failure-path per tool. Prefer runnable tests or lightweight CLI/integration checks.
Document usage for agent callers. Keep descriptions explicit so models can invoke tools correctly.

Guardrails

Keep tool names unique and verb-oriented.
Do not ship tools with ambiguous parameter names.
Every required input must be validated.
Error responses must be actionable, not generic.
If multiple transports are possible, start with stdio unless user asks otherwise.