ai-governance
Generate and enforce policy gates for AI coding agents (Copilot, Claude Code) — real-time session hooks that deny protected-path edits and dangerous commands,…
MCP server and client development — scaffold, implement tools/resources/prompts, validate schemas, debug protocol compliance, and deploy with auth and rate limiting.
> /plugin marketplace add nitinjain999/platform-skills > /plugin install platform-skills@platform-skills
How it fires
How this command gets triggered: by you, by Claude, or both.
/mcpContext preview
What this command does when you run it.
MCP server and client development — scaffold, implement tools/resources/prompts, validate schemas, debug protocol compliance, and deploy with auth and rate limiting.
name: mcp description: MCP server and client development — scaffold, implement tools/resources/prompts, validate schemas, debug protocol compliance, and deploy with auth and rate limiting. argument-hint: "[create|review|debug] [typescript|python] [description]" title: "MCP Command" sidebar_label: "mcp" custom_edit_url: null
Build, review, or debug an MCP (Model Context Protocol) server or client.
Scaffold a production-ready MCP server.
Steps: 1. Ask for: language (TypeScript or Python), transport (stdio / HTTP+SSE), list of tools and resources needed 2. Generate project scaffold with correct SDK setup 3. Implement tool handlers with Zod (TypeScript) or Pydantic (Python) schema validation 4. Add resource providers and prompt templates as needed 5. Configure transport layer 6. Add error handling — return `isError: true` content, never throw unhandled exceptions to clients 7. Add authentication and rate limiting for HTTP transports 8. Provide MCP Inspector test commands and expected responses 9. Add deployment checklist (env vars, secrets, logging)
**Validation:**
# Smoke-test the MCP server before connecting a host npx @modelcontextprotocol/inspector <your-server-command> # This opens an interactive inspector — verify: tools list, resources list, no startup errors # If the server is not discovered by the host, run /platform-skills:mcp debug
Reference: `references/mcp.md` → Protocol Fundamentals, TypeScript SDK, Python SDK
Review an existing MCP server or client implementation.
Evaluate in priority order: 1. **Protocol compliance** — JSON-RPC 2.0 correctness, proper capability negotiation, well-formed content arrays 2. **Schema validation** — all tool inputs validated with Zod/Pydantic; no `z.any()` or empty schemas 3. **Security** — no credentials in tool responses or resource content; auth on HTTP transports; rate limiting present 4. **Error handling** — errors return `isError: true` with a message, not unhandled exceptions 5. **Transport** — correct transport for the use case; no blocking sync code in async transports
Output: Critical (must fix) / Improvement (should fix) / Note (informational)
Reference: `references/mcp.md` → Security, Error Handling, Testing and Debugging
Diagnose MCP protocol or integration failures.
Classify the failure:
Evidence to collect:
# Run MCP Inspector to verify protocol compliance
npx @modelcontextprotocol/inspector node dist/index.js
# Smoke-test via stdio
echo '{"jsonrpc":"2.0","id":1,"method":"tools/list","params":{}}' | node dist/index.js
# Check for schema errors
node -e "require('./dist/index.js')" 2>&1Provide: observed symptom → likely root cause → evidence command → fix → validation step
Generate version-pinned, credential_process-backed MCP server configs for AWS services, with upfront cost warnings and CLI-first alternatives.
/platform-skills:mcp configure-aws [service|kit] [--profile <profile>] [--region <region>] [--host vscode|claude|both] [--scope global|workspace] [--multi-account]
**Examples:**
/platform-skills:mcp configure-aws eks-debug --profile prod-platform-eu --region eu-west-1 --host both /platform-skills:mcp configure-aws eks --profile dev-sandbox --host vscode --scope workspace /platform-skills:mcp configure-aws knowledge --host claude /platform-skills:mcp configure-aws eks --profile prod-eu --region eu-west-1 --multi-account
When the user invokes `configure-aws` without arguments, or omits any required value, prompt for each missing piece **one question at a time** before proceeding.
**Question 1 — Service or kit** (skip if already provided as positional arg):
Which AWS service or starter kit do you need? Services: eks EKS cluster ops, nodegroups (~2,000 tokens) cloudwatch Logs, metrics, alarms (~1,750 tokens) prometheus AMP metrics, rule groups (~1,000 tokens) iam IAM queries, read-only (~1,500 tokens) bedrock-kb Knowledge base retrieve + generate (~750 tokens) docs Real-time AWS documentation (~500 tokens) api All AWS services, catch-all (~2,500 tokens) billing Cost Explorer, budgets (~1,000 tokens) serverless SAM, Lambda lifecycle (~1,250 tokens) dynamodb Table ops, queries (~1,250 tokens) Starter kits (pre-bundled): eks-debug eks + cloudwatch (~3,750 tokens) observe cloudwatch + prometheus (~2,750 tokens) knowledge docs + bedrock-kb (~1,250 tokens) deploy api + serverless (~3,750 tokens) cost billing (~1,000 tokens) Enter service or kit name:
**Question 2 — Host** (skip if `--host` provided):
Which tool are you configuring for? 1. VS Code 2. Claude Code 3. Both Enter 1, 2, or 3:
**Question 3 — AWS profile** (skip if `--profile` provided):
Read `~/.aws/config`, list available profiles with type and environment tag (from `~/.aws-profile-tags.yaml` if present):
Which AWS profile should MCP servers use? Available profiles (from ~/.aws/config): prod-platform-eu SSO [prod] 123456789 staging-assume assumed-role [staging] 456789123 dev-sandbox SSO [dev] 987654321 Enter profile name (or press enter for no profile — uses ambient credentials):
If `~/.aws/config` is not readable
A production-grade field handbook for platform, DevOps, SRE, and cloud engineers covering Kubernetes, Flux CD, Terraform, GitHub Actions, AWS, OPA/Rego, KEDA, Karpenter, supply chain security, Falco, observability, and more.
Repo: nitinjain999/platform-skills
Generate and enforce policy gates for AI coding agents (Copilot, Claude Code) — real-time session hooks that deny protected-path edits and dangerous commands,…
Generate, convert, and maintain animated GitHub-safe Markdown documents with animated SVG diagrams. Covers four SVG patterns (architecture flow, lifecycle…
AWS profile management for MCP servers — discover profiles across SSO, Granted, and assumed-role chains, check credential TTL, switch profiles across VS Code…
Structured guidance for AWS CloudFront distributions, WAF web ACLs, Lambda@Edge, CloudFront Functions, Firewall Manager multi-account enforcement, and IAM/IRSA…
Azure identity (Workload Identity, OIDC, Entra ID), resource tagging, AKS platform patterns, RBAC scoping, and production-readiness review — with Terraform…
Design, run, and debug Chaos Engineering experiments on Kubernetes using Litmus Chaos v3 and Chaos Mesh v2. Covers fault injection (pod-delete, network-loss,…