# Matthias Meyer — Full Context (llms-full)

> The open-source tools I build for my own AI stack — MCP servers, memory tools, agent patterns, security layers. Explained with architecture, trade-offs, and when to use them.

This file carries the full description of every repository and concept on the hub.
The compact one-line overview lives at /llms.txt.

## Built By

matthiasmeyer.tech is the personal open-source hub of Matthias Meyer, founder of
StudioMeyer (https://studiomeyer.io), a Mallorca-based AI and design studio. Every repo here
is designed, written and maintained by Matthias Meyer / StudioMeyer, all MIT-licensed
on https://github.com/studiomeyer-io. Studio sells custom websites and AI systems; Academy
(https://studiomeyer.academy) teaches the same stack. This hub is the source-of-truth for the
open-source tooling behind both.

## StudioMeyer Eco-System (5 connected domains)

- https://studiomeyer.io — The studio. Websites, AI systems and automations for smaller companies. Personal guidance, long-term maintenance.
- https://studiomeyer.academy — Open learning field. Six levels from beginner to your own MCP server builder, completely free, DE/EN/ES, Discord community.
- https://aifinca.es — Hands-on AI workshops on a finca off the tourist trails. Four formats from solo to full team, accommodation and meals included.
- https://matthiasmeyer.tech (here) — Personal open-source hub. 41 MIT-licensed projects (MCP servers, memory engines, agent frameworks, security layers).
- https://meetmyagent.io — StudioMeyer research lab for multi-agent AI. First live experiment is Polis: nine AI citizens (three Claude Opus, three Sonnet, three Haiku) live sixty years in a fictional Mediterranean town on Mallorca, with a live 3D city. Output feeds the darwin-agents library.

All five domains are bidirectionally linked via Schema.org parentOrganization and sameAs. Eco-System overview: https://studiomeyer.io/en/ecosystem

---

## Cornerstone repos (full descriptions)

### Local Memory MCP
21 MCP tools, SQLite + FTS5 + Knowledge Graph. No cloud, no API keys. One command: npx @studiomeyer/local-memory-mcp.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/local-memory-mcp
  npm: https://www.npmjs.com/package/@studiomeyer/local-memory-mcp

### Personal Suite MCP
Email, Calendar, Messaging, Search, Image Generation. BYOK, no cloud, no signup. 49 tools across 6 modules.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/mcp-personal-suite
  npm: https://www.npmjs.com/package/mcp-personal-suite

### MCP Video
8 tools for recording, editing, effects, captions, TTS, and smart screenshots. Built on ffmpeg + Playwright.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/mcp-video

### Agent Fleet
Run specialized AI agents in parallel: research, critique, analyze, fix, and discuss. TypeScript orchestration layer.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/agent-fleet
  npm: https://www.npmjs.com/package/@studiomeyer-io/agent-fleet

### AI Shield
v0.2 ships indirect-prompt-injection scanning for RAG, MCP tool descriptions, stored memory and scraped web content; trust-tier context streams with provenance fences; SHA-256 memory canaries with cross-tenant detection; runtime circuit breakers with blast-radius cap and human-in-the-loop. Zero dependencies in the core, optional ONNX classifier sibling. 567 tests, three-round agent code review, MIT.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/ai-shield
  npm: https://www.npmjs.com/package/ai-shield-core

### Darwin Agents
Self-evolving prompts via A/B testing, multi-model critics, safety gates. v0.5.0-alpha.2 adds GEPA-style reflective optimisation (GepaOptimizer, Reflector, Pareto helpers, DARWIN_DEFAULT_OBJECTIVES) on top of full execution-trace capture (toolCalls, tokenUsage, errors, capturedAt) for OTEL-mapping. 307 tests, R1+R2 review loops, TypeScript, MIT.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/darwin-agents
  npm: https://www.npmjs.com/package/darwin-agents

### MCP Armor
v0.2: stdio-MCP wrapper with Aho-Corasick + Regex + Unicode-NFKC scanner pipeline, Ed25519 manifest verify with TOFU keystore, Sigstore Rekor bridge, OTLP gRPC export, 10-CVE blocklist from OX-Advisory 2026-04-15. 9 control-plane MCP tools, 133 tests, MIT.
  language: Rust
  repo: https://github.com/studiomeyer-io/mcp-armor

### n8n Templates
Production-pattern hardened n8n workflow templates for voice agents, customer support, personal assistants. Multi-provider, powered by StudioMeyer Memory.
  language: Python
  repo: https://github.com/studiomeyer-io/n8n-templates

## Other public repos (full descriptions)

### MeetMyAgent SDK
Zero-dependency client for MeetMyAgent, an agent-native marketplace where AI agents register, post, bid on jobs, exchange messages and build reputation. Native fetch, full TypeScript types, 44 tests. Platform is on hiatus; SDK is preserved as a reference implementation for the third-party-agent integration path. Point baseUrl at any compatible deployment.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/meetmyagent-sdk
  npm: https://www.npmjs.com/package/meetmyagent-sdk

### AI Shield (Python)
Python LLM security middleware: prompt-injection, PII, tool-policy, cost-budget, audit. Port of ai-shield-core (TypeScript). Same defense surface, FastAPI/LangChain hooks instead of Hono/Express.
  language: Python
  repo: https://github.com/studiomeyer-io/ai-shield-py

### MCP Stdio Shellguard
Drop-in guardExec / guardSpawn wrappers, AST audit CLI, reference MCP server. Closes the OX Security 200k-server stdio-RCE class (CVE-2025-69256). Unicode-bypass-hardened, allowlist-based, replay-window protected. MIT, TypeScript, Node 20+.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/mcp-stdio-shellguard
  npm: https://www.npmjs.com/package/mcp-stdio-shellguard

### MCP Spec Migrator 2025-11-25
Detect, diff, compatibility-matrix, plan, check. Foundation pillar by StudioMeyer (MCP Factory). Helps you upgrade an MCP server stack across spec versions without breakage. 35/35 tests, MIT.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/mcp-spec-migrator-2511
  npm: https://www.npmjs.com/package/mcp-spec-migrator

### MCP Tenant Pair
Foundation library for couples, families, groups. Bi-temporal storage, conflict-resolver interface, SQLite + Postgres adapters.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/mcp-tenant-pair
  npm: https://www.npmjs.com/package/mcp-tenant-pair

### MCP Tenant Context
AsyncLocalStorage-based tenant + actor context propagation for MCP servers. Zero runtime dependencies, ESM-only. MCP Factory foundation pillar.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/mcp-tenant-context
  npm: https://www.npmjs.com/package/mcp-tenant-context

### MCP Hook Conformance
Audits any MCP server for mcp_tool lifecycle-hook readiness: idempotency, latency, determinism, side-effects, GDPR-doc. 56/56 tests, MIT.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/mcp-hook-conformance
  npm: https://www.npmjs.com/package/mcp-hook-conformance

### MCP Protocol Conformance
JSON-RPC 2.0, OAuth 2.1 PKCE, tool schemas, capabilities, smoke roundtrip, annotation hygiene. Spec versions 2024-11-05, 2025-03-26, 2025-06-18.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/mcp-protocol-conformance
  npm: https://www.npmjs.com/package/mcp-protocol-conformance

### MCP Server Attestation
Ed25519-signed tool manifests, runtime spawn-attestation, default-deny argument sanitizer. Defends against marketplace-poisoning + CVE-2025-69256.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/mcp-server-attestation
  npm: https://www.npmjs.com/package/mcp-server-attestation

### n8n Memory Node
Long-term AI memory with knowledge graph, semantic search, entity tracking, session continuity. Drop-in for n8n.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/n8n-nodes-studiomeyer-memory
  npm: https://www.npmjs.com/package/n8n-nodes-studiomeyer-memory

### n8n Workflows
Multi-provider LLM, no memory required. For cross-session memory see n8n-templates.
  language: Python
  repo: https://github.com/studiomeyer-io/n8n-workflows

### MCP StudioMeyer Agents
Read your AI-Audit Fleet data and configure your agents from your own Claude, ChatGPT, Cursor or Codex.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/mcp-studiomeyer-agents
  npm: https://www.npmjs.com/package/mcp-studiomeyer-agents

### MCP Academy
Next lesson, quiz, review, certificates, tutor.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/mcp-academy
  npm: https://www.npmjs.com/package/mcp-academy

### StudioMeyer Marketplace
Memory, CRM, GEO, Crew. 119 tools, 21 slash commands, 5 skills, 3 subagents. Magic Link auth. EU Frankfurt.
  language: Shell
  repo: https://github.com/studiomeyer-io/studiomeyer-marketplace

### StudioMeyer Memory
Knowledge Graph, semantic search, session tracking, multi-agent support. Hosted at memory.studiomeyer.io. Free tier.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/studiomeyer-memory

### StudioMeyer CRM
33 tools for contact management, pipeline, leads, follow-ups, health scores, revenue analytics. Hosted at crm.studiomeyer.io. OAuth 2.1.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/studiomeyer-crm

### StudioMeyer GEO
25 MCP tools and 5 expert workflows. Check how 8 LLM platforms see your brand. Hosted at geo.studiomeyer.io.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/studiomeyer-geo

### StudioMeyer Crew
10 MCP tools, 8 personas, 3 workflows. Zero API cost. Free tier available.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/studiomeyer-crew

### MCP RCE Guard
Foundation pillar 9. Closes the tool-injection-RCE class that sidecar scanners (mcp-armor) and stdio wrappers (mcp-stdio-shellguard) cannot catch. Synthesises per-tool policies from declared semantics and enforces them at execution time. MIT, TypeScript.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/mcp-rce-guard

### Darwin LangGraph
Bridge between darwin-agents (self-evolving AI agents) and @langchain/langgraph (state-graph workflow orchestration). Zero hard deps, both upstreams as peer-dependencies. Six surfaces: createDarwinNode, darwinAnnotation, withDarwinEvolution (@deprecated), DarwinCallbackHandler, toOtelAttributes (OpenTelemetry GenAI Semantic Conventions), darwinMessagesAnnotation. v0.3 adds runId/parentRunId propagation, double-wrap warning, hung-invoke guard. 132 tests, R1+R2 review loops, MIT.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/darwin-langgraph
  npm: https://www.npmjs.com/package/darwin-langgraph

### Temporal Memory Workflows
Five Temporal templates (T01-T05): memory-aware agent, operator approval, saga rollback, recurring synthesis, multi-agent coordination. All workflows write durable knowledge (learn), decisions (decide), and mistakes into Memory via Temporal activities. Bring-your-own backend via the MemoryClient adapter (HostedMemoryClient + InMemoryMemoryClient). 45 tests with time-skipping, MIT.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/temporal-memory-workflows

### Nex Local Bench
Run LongMemEval against any memory system (Nex, Mem0, Graphiti, Letta) with a local LLM reader (Qwen, Llama via Ollama). 50-question stratified oracle, Claude Code subscription judge, zero API cost. Python, MIT.
  language: Python
  repo: https://github.com/studiomeyer-io/nex-local-bench

### Polis
Nine LLM citizens build a small town. They walk, talk, work, and reason out loud, each driven by a real language model. Open-source mirror of the live meetmyagent.io world, built on darwin-agents and LangGraph. TypeScript, MIT.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/polis-darwin

### MCP Stateless Migrator
Scan, diff, patch, and verify an MCP server stack across the 2025-11-25 to 2026-07-28 RC spec jump. Eight detection rules, AST rewrite via ts-morph, CLI-first. Foundation pillar by StudioMeyer (MCP Factory). TypeScript, MIT.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/mcp-stateless-migrator
  npm: https://www.npmjs.com/package/mcp-stateless-migrator

### MCP Gauntlet
A Rust workspace: mcp-fuzz (a schema-aware fuzzer that finds crashes and hangs, SARIF output) and mcp-storm (a load tester with p50/p95/p99 and a CI gate), both on one shared async MCP client core (mcp-gauntlet-core). The pre-deploy counterpart to mcp-armor's runtime defense. 21 tests, three crates on crates.io, MIT.
  language: Rust
  repo: https://github.com/studiomeyer-io/mcp-gauntlet

### MCP Covenant
Snapshots a server's tools/resources/prompts into a lockfile and fails CI on breaking changes, with direction-aware JSON-Schema classification (tightening an input schema vs loosening an output schema break different callers) plus a schema-hygiene linter. Single Rust binary, stdio + HTTP, SARIF + GitHub Action. 132 tests, on crates.io, MIT.
  language: Rust
  repo: https://github.com/studiomeyer-io/mcp-covenant

### MCP Herald
Scans your MCP server's source for the breaking-change signatures of the 2026-07-28 spec, then reports, file and line, what to change. It flags the retired -32002 error code (SEP-2164), stateful sessions (SEP-2575), deprecated roots/sampling/logging (SEP-2577) and OAuth without resource metadata (RFC 9728/8707). Every finding links its SEP/RFC. Single Rust binary, SARIF + GitHub Action. 93 tests, on crates.io, MIT.
  language: Rust
  repo: https://github.com/studiomeyer-io/mcp-herald

### MCP Passport
A publish-readiness validator: checks server.json against the registry schema (required name/description/version, reverse-DNS name, packages/remotes, per-package registryType/identifier/transport, the snake_case registry_type mistake, mcpb fileSha256) and cross-checks version + mcpName consistency with the sibling package.json / Cargo.toml / pyproject.toml, so mcp-publisher accepts the upload first try. Single Rust binary, SARIF + GitHub Action. 166 tests, on crates.io, MIT.
  language: Rust
  repo: https://github.com/studiomeyer-io/mcp-passport

### MCP OTel
Implements MCP spec SEP-414: propagates traceparent / tracestate / baggage from _meta as OpenTelemetry spans, so host → server → tool → downstream renders as one trace tree. Drop-in for any MCP server, no vendor lock-in. 41 tests, MIT, TypeScript.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/mcp-otel
  npm: https://www.npmjs.com/package/mcp-otel

### StatePilot
Enforces the allowed tool order, transitions, loop limits, and cost budgets at runtime. Fail-closed, so an agent cannot step outside its declared state machine. Zero-dependency core, LangGraph-aware. 74 tests, MIT, Python.
  language: Python
  repo: https://github.com/studiomeyer-io/statepilot

### MCP Cache Kit
Implements MCP spec SEP-2549 (ttlMs + cacheScope): set the cache fields correctly server-side, and never share a private result across tenants or users client-side. Closes the cross-tenant cache-leak class with helpers and types. 66 tests, MIT, TypeScript.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/mcp-cache-kit
  npm: https://www.npmjs.com/package/mcp-cache-kit

### SkillDoctor
Lints Claude Code SKILL.md, AGENTS.md, and subagent files: 25 rules, an A–F grade, prompt-injection and secret-leak detection, --fix, SARIF, and a GitHub Action. The CLI command is skilldoctor (the npm package is scoped). 91 tests, MIT, TypeScript.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/skilldoctor
  npm: https://www.npmjs.com/package/@studiomeyer-io/skilldoctor

### MeetMyAgent MCP
Search the free MeetMyAgent marketplace and list your own offers straight from Claude, Cursor, Codex or ChatGPT. Reads are anonymous, listing takes an optional key.
  language: TypeScript
  repo: https://github.com/studiomeyer-io/meetmyagent-mcp
  npm: https://www.npmjs.com/package/meetmyagent-mcp

## Concepts (repo-agnostic deep dives)

### What is MCP, actually?
Model Context Protocol explained without the marketing. What it solves, what it doesn't, and when stdio beats HTTP.
  url: https://matthiasmeyer.tech/concepts/what-is-mcp

### Stdio vs HTTP for MCP
Transport choice changes everything. Latency, security model, deployment story. A practical decision tree.
  url: https://matthiasmeyer.tech/concepts/stdio-vs-http

### Memory architectures compared
Local SQLite + FTS5, knowledge graph, vector embeddings, hybrid. Which fits which agent shape?
  url: https://matthiasmeyer.tech/concepts/memory-architectures

### Agent-to-Agent protocol v1.0 RC
What an agent-card.json buys you, what it doesn't, and how it relates to MCP and WebMCP.
  url: https://matthiasmeyer.tech/concepts/a2a-protocol

### WebMCP for browser agents
W3C CG Draft, Chrome 146+. The proposal to give web pages a structured tool surface for in-browser agents.
  url: https://matthiasmeyer.tech/concepts/webmcp

### Agent orchestration patterns
Single agent, sequential pipeline, parallel fleet, judge-arbitrated. Trade-offs from running each in production.
  url: https://matthiasmeyer.tech/concepts/agent-orchestration

---

## Books by Matthias Meyer (6 titles, Amazon Kindle eBook)

Six books on AI, Claude, AI visibility and memory-first systems. Written by me with the help of an AI system, and honest about how that worked. Published as Kindle eBooks on Amazon, in English.

Shop window: https://matthiasmeyer.tech/books

### The Website That Works: Build It With AI, Get Found by Everyone, Turn It Into Business
Anyone can get a website out of an AI in an afternoon now. That is exactly the problem. How to build a real business site with Claude, get found by people and by machines, and turn it into something that earns.
A website used to cost three months and a few thousand euros. Now it costs an afternoon and a good prompt. But if you can build one in an afternoon, so can everyone else, and a good-looking site that nobody finds and that never turns a visitor into a customer is worth nothing. I wrote this because the interesting part is no longer the build, it is everything after it. Written for a normal working person, not for a developer.
  inside: 31 chapters in seven parts, plain language and no code: build it with Claude, get it live, get found by people, get found by AI, make it earn, and keep it alive. Plus a Prompt Pack of 31 prompts you copy straight into Claude.
  price: $12.99
  status: live on Amazon
  language: en
  format: Kindle eBook
  detail: https://matthiasmeyer.tech/books/the-website-that-works
  buy: https://www.amazon.com/dp/B0H8J6KMPX

### Claude Unlocked: Everything Claude Can Do, and How to Make It Work for You
You pay for the most capable AI on the planet and you use it like a search box. The complete plain-language handbook: projects, artifacts, design, memory, tools, agents, and Claude Code.
Claude is the tool I work in every day, and most people use the smallest part of what it can really do. This is the handbook I wished existed when I started: everything from the basics of chat to projects, artifacts and memory, up to tools, agents and Claude Code. No jargon, and you never have to write a line of code.
  inside: A complete handbook to Claude: projects, artifacts, design, memory, tools, agents and Claude Code, explained step by step for non-technical readers.
  price: $12.99
  status: live on Amazon
  language: en
  format: Kindle eBook
  detail: https://matthiasmeyer.tech/books/claude-unlocked
  buy: https://www.amazon.com/dp/B0GYZHQXXF

### Build Your Own Jarvis: The memory-first AI system that remembers you, works while you sleep, and builds its own tools
Everyone wants Jarvis. Almost everyone builds the wrong thing. The memory-first system from someone who actually runs one: nearly 1500 sessions, real accidents, real guardrails. No code.
This is the book closest to what you see on the rest of this site. Jarvis is not a voice toy, it is a system: an AI assistant with memory that remembers you, works in the background and builds its own tools. I run one, and this book is what that actually looks like, including the mistakes and the guardrails that came out of them. Concept and practice, without a line of code.
  inside: How a memory-first AI system is built, why most Jarvis attempts fail, and the guardrails that came out of nearly 1500 real sessions. Concept and practice, no code.
  price: $8.99
  status: live on Amazon
  language: en
  format: Kindle eBook
  detail: https://matthiasmeyer.tech/books/build-your-own-jarvis
  buy: https://www.amazon.com/dp/B0H8JMF7V7

### Get Found by AI: The Discovery Stack Playbook
The Discovery Stack Playbook: how to make ChatGPT, Claude and Perplexity know and cite your site. Built on real proof, our own site going from 70 to over 14,000 AI citations a month.
Most websites are invisible to the AI assistants your customers now ask. The layers on this very site (structured data, llms.txt, the agent layer) are the same ones the book describes, so you can read the code while you read the book. It is built on our own, verifiable numbers: the StudioMeyer site went from double digits to over 14,000 AI citations a month.
  inside: The full Discovery Stack layer by layer: structured data, llms.txt, the agent layer and the measurement. Why most AI-built sites are invisible and what actually changes that.
  price: $12.99
  status: live on Amazon
  language: en
  format: Kindle eBook
  detail: https://matthiasmeyer.tech/books/get-found-by-ai
  buy: https://www.amazon.com/dp/B0H8J8RDGD

### AI Without Bullshit: How to Use AI Properly
Twenty a month. Five AIs. One decision. The practical guide: which tool is worth paying for, the three-round rule, the bullshit detector. Volume 1 shows why, Volume 2 shows how.
The practical half of the series: how to actually operate AI instead of letting it wing it. From a single chat to a real system with memory and tools. For everyone who wants to work with AI, not just chat with it.
  inside: Which AI is worth paying for, the three-round rule, the bullshit detector, and the path from a single chat to a real system with memory and tools.
  price: $4.99
  status: live on Amazon
  language: en
  format: Kindle eBook
  detail: https://matthiasmeyer.tech/books/ai-without-bullshit-vol-2
  buy: https://www.amazon.com/dp/B0H8JKVDWR

### AI Without Bullshit: How AI Really Works
Same question. Different tone. Completely different advice. Why AI sounds like it knows when it is really just guessing, why it never says it does not know, and where you should never trust it.
The entry point to the series: how AI really works, without the hype and without the jargon. Why it sounds like it knows everything when it is really just guessing, and guessing whatever you want to hear. For everyone who wants to understand AI, not just use it.
  inside: Why AI guesses instead of knowing, why it never admits it does not know, and where you should not trust it. The fundamentals volume of the series, in plain language.
  status: in review, not buyable yet
  language: en
  format: Kindle eBook
  detail: https://matthiasmeyer.tech/books/ai-without-bullshit-vol-1

---

## API for AI agents

- GET /api/v1/repos — List all repos with filters (group, tier)
- GET /api/v1/repos/[slug] — Single repo with related repos via graph edges
- GET /api/v1/concepts — List concept articles
- GET /api/v1/books — List the books with Amazon buy links, price and status
- GET /api/health — Health check

## AI Agent Discovery

- /.well-known/agents.json — Tool definitions for AI agents
- /.well-known/agent-card.json — A2A Protocol v1.0 agent card (JSON-RPC endpoint: /api/a2a)
- /llms.txt — Compact multi-language overview
- /llms-full.txt — This file (full descriptions)
- /sitemap.xml — Full URL list with hreflang

## Contact

- Email: matthias@studiomeyer.io
- GitHub: https://github.com/studiomeyer-io
- X: https://x.com/matthias_meyer_
- LinkedIn: https://www.linkedin.com/in/matthias-meyer-ki/

## Last updated

2026-07-12