--- title: "Recipes: from CLAUDE.md to your own MCP server" description: "58 step-by-step recipes, from CLAUDE.md basics to building, deploying and distributing your own MCP server. Free to read, no account needed." canonical: "https://studiomeyer.academy/en/recipes" language: en image: "https://studiomeyer.academy/api/og/page/share?title=Recipes%3A+from+CLAUDE.md+to+your+own+MCP+server§ion=Recipes&locale=en" markdown_versions: ["https://studiomeyer.academy/en/recipes.md"] publisher: "StudioMeyer Academy, https://studiomeyer.academy (llms.txt: https://studiomeyer.academy/llms.txt)" --- Recipes · AI Operator Stack # 58 fertige Recipes, *vom CLAUDE.md zum Multi-Tenant-MCP* Memory-First Operator-Stack als kuratierter Pfad. 58 Recipes über 16 Phasen. Kein Theorie-Talk: jede Recipe endet damit dass auf Deinem System etwas konkret läuft. ## Dein Fortschritt 01. Foundation0/4 02. Connectivity0/5 03. Basic Workflows0/7 04. Memory & Knowledge0/10 05. Daily Use0/5 06. Build MCP0/5 07. MCP Patterns0/2 08. Deploy MCP0/5 09. Distribution0/6 10. Multi-Tenant SaaS0/3 11. Operator Skills0/1 16. Phase 160/5 ## Schnellfilter [Alle (58)](https://studiomeyer.academy/recipes) [Phase 1 (4)](https://studiomeyer.academy/recipes?phase=1) [Phase 2 (5)](https://studiomeyer.academy/recipes?phase=2) [Phase 3 (7)](https://studiomeyer.academy/recipes?phase=3) [Phase 4 (10)](https://studiomeyer.academy/recipes?phase=4) [Phase 5 (5)](https://studiomeyer.academy/recipes?phase=5) [Phase 6 (5)](https://studiomeyer.academy/recipes?phase=6) [Phase 7 (2)](https://studiomeyer.academy/recipes?phase=7) [Phase 8 (5)](https://studiomeyer.academy/recipes?phase=8) [Phase 9 (6)](https://studiomeyer.academy/recipes?phase=9) [Phase 10 (3)](https://studiomeyer.academy/recipes?phase=10) [Phase 11 (1)](https://studiomeyer.academy/recipes?phase=11) [Phase 16 (5)](https://studiomeyer.academy/recipes?phase=16) ## Foundation FreePhase 1–5 Phase 1 · 01 FreeOpen ### CLAUDE.md / AGENTS.md, give Claude memory across sessions Two markdown files Claude reads on every session start. The single biggest leverage point in your setup. 5 steps https://studiomeyer.academy/recipes/1.1-claude-md Phase 1 · 02 FreeOpen ### Skills, package recurring workflows so Claude triggers them automatically Skills are markdown files Claude loads on demand based on natural-language triggers. Like slash commands but smarter. 5 steps https://studiomeyer.academy/recipes/1.2-skills Phase 1 · 03 FreeOpen ### Hooks, turn Claude's behavior into automated guardrails Hooks run shell commands on Claude Code events (before tool use, after edit, on session start). They're the difference between 'Claude does X' and 'Claude can never do Y'. 6 steps https://studiomeyer.academy/recipes/1.3-hooks Phase 1 · 04 FreeOpen ### settings.json + Permissions, stop the permission popup spam By default Claude Code asks before running any new shell command. Configure permission scopes once and the right things become auto-allowed forever. 6 steps https://studiomeyer.academy/recipes/1.4-settings-permissions Phase 2 · 01 FreeOpen ### Your first MCP server, claude mcp add Install your first MCP server in Claude Code or Codex in under two minutes. We pick context7 (library docs), small, useful, no API key. 5 steps https://studiomeyer.academy/recipes/2.1-first-mcp Phase 2 · 02 FreeOpen ### Memory layer, give your AI persistent memory across sessions Claude Code's built-in memory is per-project. Codex has none. Adding mcp-nex (or local-memory-mcp) gives you durable, searchable memory that works across every MCP client you use. 4 steps https://studiomeyer.academy/recipes/2.2-memory-layer Phase 2 · 03 FreeOpen ### Web research. Brave, Exa, and SearXNG side-by-side Three MCP servers that cover 95% of research needs. Brave for fast keyword search, Exa for semantic / neural, SearXNG via mcp-searxng for unlimited free queries. 5 steps https://studiomeyer.academy/recipes/2.3-web-research Phase 2 · 04 FreeOpen ### Code intelligence, codebase-memory-mcp + codegraph Index your repo into a knowledge graph so Claude can answer 'who calls this function' and 'what breaks if I delete it' in one tool call instead of grepping. 6 steps https://studiomeyer.academy/recipes/2.4-code-intelligence Phase 2 · 05 FreeOpen ### GitHub MCP, read PRs, issues, code from any repo Official GitHub MCP server gives Claude direct access to PRs, issues, code search, and review workflows. Hosted remote (api.githubcopilot.com/mcp) or local Docker. 19 toolsets. 5 steps https://studiomeyer.academy/recipes/2.5-github-mcp Phase 3 · 01 FreeOpen ### Slash commands and built-in shortcuts /model, /effort, /context, /memory, /agents, the built-in commands that change how the session behaves. Plus how to write your own. 3 steps https://studiomeyer.academy/recipes/3.1-slash-commands Phase 3 · 02 FreeOpen ### Permissions reloaded, building a tight allow-list Beyond the safe defaults from Recipe 1.4: a project-scoped allow-list that pre-clears 90% of routine prompts without giving up the dangerous-command guardrails. 5 steps https://studiomeyer.academy/recipes/3.2-permissions-plus Phase 3 · 03 FreeOpen ### Subagents 101, when to delegate vs. when to do it yourself Custom subagents in ~/.claude/agents/. Frontmatter, isolation, when an agent beats a skill, when it does not. 5 steps https://studiomeyer.academy/recipes/3.3-subagent-basics Phase 3 · 04 FreeOpen ### Project layout. CLAUDE.md, .claude/, .mcp.json conventions Where to put project-scoped config so Claude Code, Cursor, and Codex all read the same file. Plus the .mcp.json convention for sharing MCP servers with a team. 5 steps https://studiomeyer.academy/recipes/3.4-project-layout Phase 3 · 05 FreeOpen ### Multi-tool setup. Claude Code + Codex + Cursor on one machine Sync MCP servers between three clients with one config script. Avoid port collisions and config drift. 5 steps https://studiomeyer.academy/recipes/3.5-multi-tool-setup Phase 3 · 06 FreeOpen ### Branch-per-Ask, give every AI task its own sandbox Stop letting Claude or Cursor experiment on your working state. One branch per meaningful AI task. Throwaway when bad, merge when good. 6 steps https://studiomeyer.academy/recipes/3.6-branch-per-ask Phase 3 · 07 FreeOpen ### Safe AI rollback, never let Claude run git reset --hard for you Three rollback paths that do not destroy uncommitted work. Plus the one Claude Code bug you need to know about so you stop saying 'undo it' to the AI. 6 steps https://studiomeyer.academy/recipes/3.7-safe-ai-rollback Phase 4 · 01 FreeOpen ### Persistent memory, make your AI use it without being asked Connect a memory MCP (StudioMeyer Memory, local-memory-mcp, Mem0, pick one) and flip the always-on switch. After 5 minutes your assistant searches memory before answering and saves insights automatically. 5 steps https://studiomeyer.academy/recipes/4.1-persistent-memory Phase 4 · 02 FreeOpen ### Project tags, keeping multi-project memory clean Tag every memory with a project slug. Filter searches by project. Avoid the 'all my projects bleed into each other' failure mode. 6 steps https://studiomeyer.academy/recipes/4.2-project-tags Phase 4 · 03 FreeOpen ### Knowledge graph, entities, observations, relationships Beyond flat learnings: nex\_entity\_create, nex\_entity\_observe, nex\_entity\_relate. When the graph beats search. 5 steps https://studiomeyer.academy/recipes/4.3-knowledge-graph Phase 4 · 04 FreeOpen ### Importing ChatGPT and Claude.ai history Export your existing conversations and pipe them into nex\_import. Years of context become searchable memory in one batch. 6 steps https://studiomeyer.academy/recipes/4.4-chatgpt-import Phase 4 · 05 FreeOpen ### Cross-tool memory, same memory in Claude Code, Codex, Cursor Point all three clients at the same memory MCP. Discuss tradeoffs of local-stdio vs SaaS for cross-device sharing. 7 steps https://studiomeyer.academy/recipes/4.5-cross-tool-memory Phase 4 · 06 FreeOpen ### CLAUDE.md or memory server, pick the right tool A six-question decision tree that tells you whether a markdown file is enough or whether you need a structured memory server. Most projects do not need a server. 7 steps https://studiomeyer.academy/recipes/4.6-claude-md-vs-memory Phase 4 · 07 FreeOpen ### Auto-Memory and Auto-Dream, what Claude Code does at night Anthropic added Auto-Memory in March 2026 and Auto-Dream consolidates it nightly. Useful by default. Dangerous when combined with external memory systems. Here is how to keep both happy. 5 steps https://studiomeyer.academy/recipes/4.7-auto-memory-and-auto-dream Phase 4 · 08 FreeOpen ### Memory server comparison. Mem0 vs Zep vs Letta vs alternatives Honest side-by-side of the major MCP memory servers in 2026. Pricing that hides knowledge graph behind premium, benchmark scores nobody mentions, and the LongMemEval numbers. 5 steps https://studiomeyer.academy/recipes/4.8-memory-server-comparison Phase 4 · 09 FreeOpen ### The risks no memory vendor mentions, sycophancy, stale facts, privacy drift Memory amplifies sycophancy. It accumulates stale facts. It leaks context across topics. None of the major vendors solve this. Here is what to do about it. 5 steps https://studiomeyer.academy/recipes/4.9-sycophancy-and-stale-facts Phase 4 · 10 FreeOpen ### Memory hygiene, the weekly, monthly, quarterly routine Memory without care degrades. Here is the maintenance schedule that keeps it useful for years instead of months. Five tasks per week, three per month, one per quarter. Total time: 20 minutes a week. 5 steps https://studiomeyer.academy/recipes/4.10-memory-hygiene-routine Phase 5 · 01 FreeOpen ### Morning brief, start the day with a cross-domain status Sprint plus system health plus open follow-ups plus failed jobs in one prompt. Install the morning-brief skill or write your own. 6 steps https://studiomeyer.academy/recipes/5.1-morning-brief Phase 5 · 02 FreeOpen ### /schedule and recurring agents Claude Code's /schedule slash command creates cloud-managed Routines that run on a recurring cadence, on demand via API, or on GitHub events. When to use it, what it actually does, and the gotchas we hit. 8 steps https://studiomeyer.academy/recipes/5.2-schedule-routines Phase 5 · 03 FreeOpen ### Debrief, clean session ends save next-session context End of session ritual: summarize done / open / blocked / decisions. The /debrief slash command plus optional Stop hook so it runs automatically. 7 steps https://studiomeyer.academy/recipes/5.3-debrief Phase 5 · 04 FreeOpen ### Notifications. Telegram, Discord, email signals from your AI When a long-running agent finishes you want to know. Wire up mcp-multi-channel or an email hook so signals reach you outside the terminal. 8 steps https://studiomeyer.academy/recipes/5.4-notifications Phase 5 · 05 FreeOpen ### Weekly review, synthesize a week of memory into one summary nex\_synthesize, nex\_reflect, nex\_communities. Use them on Friday afternoon to surface patterns you would not see day-by-day. 7 steps https://studiomeyer.academy/recipes/5.5-weekly-review ## Build & Deploy your own MCP FreePhase 6–10 Phase 6 · 01 FreeOpen ### Build Your Own MCP, minimal server skeleton in TypeScript From npm init to a working MCP server with one tool, in 60 lines of TypeScript. The skeleton you can fork for everything in Phase 6-10. 6 steps https://studiomeyer.academy/recipes/6.1-server-skeleton Phase 6 · 02 FreeOpen ### Tool design, what makes a good MCP tool Naming, descriptions, input schemas, idempotency, error responses. The difference between a tool the LLM uses correctly and one it ignores or breaks. 6 steps https://studiomeyer.academy/recipes/6.2-tool-design Phase 6 · 03 FreeOpen ### Input validation. Zod patterns that don't fight the LLM Why every MCP tool needs runtime validation, the Zod patterns that work, the strictness traps that break LLM calls, and the error responses that let the LLM recover. 5 steps https://studiomeyer.academy/recipes/6.3-input-validation Phase 6 · 04 FreeOpen ### Testing MCP tools, vitest + in-memory transport How to write tests for MCP tools without spawning subprocesses. Unit tests for handlers, integration tests via in-memory transport, smoke tests for stdio mode. 6 steps https://studiomeyer.academy/recipes/6.4-testing-mcp-tools Phase 6 · 05 FreeOpen ### Logging pattern, stderr-only, structured, never on stdout The single most common production bug in homemade MCP servers: a stray console.log corrupts the wire protocol. This recipe is the structured-logging pattern that prevents it. 5 steps https://studiomeyer.academy/recipes/6.5-logging-pattern Phase 7 · 04 FreeOpen ### Database. Postgres pool, idempotent migrations, the connection-leak trap Pool sizing for serverless vs long-lived servers, ssl rejectUnauthorized for Supabase, idempotent DDL via DO blocks, and the connection-leak pattern that takes down a SaaS silently. 7 steps https://studiomeyer.academy/recipes/7.4-database-pattern Phase 7 · 05 FreeOpen ### Error handling, what to surface, what to swallow isError responses that let the LLM recover, never leak stack traces, never 500 on user-controllable input. The error-shape contract that keeps your support inbox empty. 7 steps https://studiomeyer.academy/recipes/7.5-error-handling Phase 8 · 01 FreeOpen ### Deploy to MCPize Cloud Run, managed hosting in 5 minutes MCPize is the easiest path from npm package to hosted MCP server. Container builds via Buildpacks, Cloud Run runs it, you get an HTTPS URL. The PORT trigger gotcha + Prisma custom-output trap. 9 steps https://studiomeyer.academy/recipes/8.1-mcpize-cloud-run Phase 8 · 02 FreeOpen ### Docker Compose, direct hosting on a VPS you own Self-host on Hetzner / DigitalOcean / your own server. node:22-slim base, env\_file pattern, health check, the --force-recreate gotcha. 7 steps https://studiomeyer.academy/recipes/8.2-docker-compose Phase 8 · 03 FreeOpen ### Nginx + Let's Encrypt, production TLS in 10 lines DNS-01 wildcard certs via Cloudflare, HSTS + security headers, the 3600s SSE timeout, the Cloudflare real-IP setup. 8 steps https://studiomeyer.academy/recipes/8.3-nginx-proxy Phase 8 · 04 FreeOpen ### /health endpoint, liveness vs readiness, the 25h-down lesson What goes in /health, what stays out, why a DB-roundtrip-less /health is treacherous, and the one-liner that prevents your container from looking healthy while the DB is down. 8 steps https://studiomeyer.academy/recipes/8.4-health-endpoint Phase 8 · 05 FreeOpen ### Blue-green deploys, zero-downtime container swaps with nginx Two containers behind one nginx upstream, atomic switch via reload, parallel run for soak time, kill the old. Production deploys without dropping a single request. 7 steps https://studiomeyer.academy/recipes/8.5-blue-green Phase 9 · 01 FreeOpen ### Publish to npm, bin field, files allowlist, npx -y Make your server installable via 'claude mcp add foo -s user -- npx -y your-package'. The four package.json fields that decide whether it works for users. 7 steps https://studiomeyer.academy/recipes/9.1-npm-publish Phase 9 · 02 FreeOpen ### GitHub OSS, trust signals that get you stars README structure, SECURITY.md, CI badge, OpenSSF Scorecard, npm provenance via OIDC, Dependabot. The trust stack that gets your repo past skeptical reviewers. 9 steps https://studiomeyer.academy/recipes/9.2-github-oss Phase 9 · 03 FreeOpen ### MCP marketplaces, where to submit, what each one auto-pulls mcp.so, MCPize, awesome-mcp-servers, Smithery, Glama. Submission process, what each directory auto-pulls from your README, how to spot fake-traffic ones. 9 steps https://studiomeyer.academy/recipes/9.3-mcp-marketplaces Phase 9 · 04 FreeOpen ### Trusted publishing, token-less npm releases via GitHub OIDC Replace NPM\_TOKEN secrets with OIDC. publish.yml that triggers on tag push. npm provenance badges. Cleaner trust signal for security-conscious users. 8 steps https://studiomeyer.academy/recipes/9.4-trusted-publishing Phase 9 · 05 FreeOpen ### Writing a README that actually gets installed First three lines decide. Show the install command, the 'what it does' in one sentence, and a copy-paste config snippet. The README anatomy that turns viewers into users. 10 steps https://studiomeyer.academy/recipes/9.5-readme-that-converts Phase 9 · 06 FreeOpen ### Breaking changes, evolving a published server without breaking users A tool definition has no version field and no deprecated flag. So the package is your only version signal, and most users are pinned to nothing. The alias window, the semver rules, and the snapshot test that catches an accidental rename before publish. 10 steps https://studiomeyer.academy/recipes/9.6-breaking-changes Phase 10 · 01 FreeOpen ### Stripe Checkout, subscription mode end-to-end From Stripe dashboard product to working checkout in your MCP SaaS: setup script, /api/billing/checkout endpoint, success\_url handling, and the customer portal for self-service. 7 steps https://studiomeyer.academy/recipes/10.1-stripe-checkout Phase 10 · 02 FreeOpen ### Magic-link mail. Brevo SMTP, sender reputation, the +tag trap Passwordless auth via one-time email links. The DKIM/SPF/DMARC setup that prevents Yahoo bounces, the token TTL pattern, the +tag email gotcha, and the deliverability checklist. 7 steps https://studiomeyer.academy/recipes/10.2-magic-link-mail Phase 10 · 05 FreeOpen ### Stripe webhooks, idempotent, signature-verified, fail-safe HMAC-SHA256 signature verification, idempotency on event\_id, the raw-body trap, the full subscription lifecycle, and the retry behavior you need to plan for. 8 steps https://studiomeyer.academy/recipes/10.5-stripe-webhooks ## Operator-grade Production FreePhase 11–16 Phase 11 · 06 FreeOpen ### Model allowlist, lock the team to the models you actually pay for Use availableModels to restrict which models a session can pick, and enforceAvailableModels in managed settings so project or user config can't widen it. Stops surprise Opus bills and keeps a team on an approved set. 6 steps https://studiomeyer.academy/recipes/11.6-model-allowlist Phase 16 · 01 FreeOpen ### The mcp\_tool hook, direct MCP tool calls from Claude Code lifecycle events Claude Code v2.1.118 added a new hook type: type: 'mcp\_tool'. Skip the bash wrapper, call MCP server tools directly from Stop, PreCompact, UserPromptSubmit. The five-rule sanity check before you wire one. 6 steps https://studiomeyer.academy/recipes/16.1-mcp-tool-hook-intro Phase 16 · 02 FreeOpen ### Memory hook bundle, auto-persist sessions, snapshot before compact, recall on trigger Four mcp\_tool hooks for the StudioMeyer Memory MCP that turn it from 'a tool I call manually' into 'a memory layer that always knows what I did'. Stop, PreCompact, UserPromptSubmit, SubagentStop, all idempotent, all fast. 6 steps https://studiomeyer.academy/recipes/16.2-memory-hook-bundle Phase 16 · 03 FreeOpen ### CRM hook bundle, auto-lookup customers, log interactions from email drafts Two mcp\_tool hooks for the StudioMeyer CRM that turn customer-mention into instant context and email-draft into automatic interaction log. UserPromptSubmit, PostToolUse on Edit/Write. 5 steps https://studiomeyer.academy/recipes/16.3-crm-hook-bundle Phase 16 · 04 FreeOpen ### GEO + Crew hook bundles, auto-audit content edits, auto-activate persona on session start Stop hook with file-pattern filter triggers GEO audit when you edit markdown. SessionStart hook activates the right Crew persona based on cwd. Two bundles, two MCPs, both production-ready. 6 steps https://studiomeyer.academy/recipes/16.4-geo-crew-hook-bundle Phase 16 · 05 FreeOpen ### Academy hook bundle, auto-recipe lookup, session-start dashboard Two hooks for the Academy MCP itself. UserPromptSubmit points the assistant at a matching recipe when you ask 'how do I X'. SessionStart shows XP, rank and next lesson without typing a command. 6 steps https://studiomeyer.academy/recipes/16.5-academy-hook-bundle ## Sitemap Public pages of this site as Markdown: [sitemap](https://studiomeyer.academy/sitemap.md). Curated entry point: [llms.txt](https://studiomeyer.academy/llms.txt), full text: [llms-full.txt](https://studiomeyer.academy/llms-full.txt).