{"slug": "mcp-tool-stack-13-servers-180-tools-ai-native-dev-workflow", "title": "🔧 MCP Tool Stack — 13 servers, ~180 tools, AI-native dev workflow", "summary": "Model Context Protocol (MCP) tool stack, which integrates 13 specialized servers offering approximately 180 tools for an AI-native development workflow. These servers cover a wide range of functions including web intelligence, GitHub lifecycle management, AI/ML ecosystem exploration, data science competitions, ML experiment tracking, academic research, browser automation, and local file operations.", "body_md": "# 🔧 MCP Tool Stack · Enes Demir\n\n> **Model Context Protocol** — 13 servers, ~180 tools, AI-native development\n>\n> _Last updated: 2026-05-22 · Hermes Agent · DeepSeek V4 Flash_\n\n---\n\n## 📡 The Stack\n\n| # | Server | Tools | Area | Link |\n|---|--------|-------|------|------|\n| 1 | **Firecrawl** | ~24 | Web scraping, search, crawl, extraction, monitoring | [firecrawl.dev](https://firecrawl.dev) |\n| 2 | **GitHub** | ~30+ | PRs, issues, repos, code review, CI, releases | [MCP GitHub Server](https://github.com/modelcontextprotocol/servers/tree/main/src/github) |\n| 3 | **HuggingFace** | ~20 | Hub search, Spaces, papers, datasets, HF Jobs | [huggingface.co/mcp](https://huggingface.co/docs/hub/en/mcp) |\n| 4 | **Kaggle** | ~50+ | Competitions, datasets, models, notebooks, leaderboards | [kaggle.com/docs/mcp](https://www.kaggle.com/docs/mcp) |\n| 5 | **W&B (Weights & Biases)** | ~15 | Run tracking, traces, artifacts, reports, evals | [github.com/wandb/wandb-mcp-server](https://github.com/wandb/wandb-mcp-server) |\n| 6 | **arXiv** | ~12 | Paper search, download, citation graph, alerts | [github.com/blazickjp/arxiv-mcp-server](https://github.com/blazickjp/arxiv-mcp-server) |\n| 7 | **Browser Use** | ~13 | Headless browser automation, screenshots, interaction | [github.com/browser-use/browser-use](https://github.com/browser-use/browser-use) |\n| 8 | **Filesystem** | ~13 | File read/write, search, directory tree, metadata | [MCP Filesystem Server](https://github.com/modelcontextprotocol/servers/tree/main/src/filesystem) |\n| 9 | **Context7** | ~2 | Library documentation query (Next.js, React, etc.) | [mcp.context7.com](https://mcp.context7.com) |\n| 10 | **Git** | ~10 | Local git operations (diff, log, branch, commit) | [mcp-server-git](https://github.com/modelcontextprotocol/servers/tree/main/src/git) |\n| 11 | **Time** | ~2 | Timezone conversion, current time | [MCP Time Server](https://github.com/modelcontextprotocol/servers/tree/main/src/time) |\n| 12 | **Sequential Thinking** | ~1 | Structured multi-step reasoning | [@modelcontextprotocol/server-sequential-thinking](https://github.com/modelcontextprotocol/servers/tree/main/src/sequentialthinking) |\n| 13 | **Supabase** | ~? | Database, auth, storage via MCP | [supabase.com/docs/guides/platform/mcp](https://supabase.com/docs/guides/platform/mcp) |\n\n---\n\n### 🔥 Firecrawl\n```\ncrawl  →  scrape  →  search  →  map  →  extract  →  agent\n```\n[🌐 firecrawl.dev](https://firecrawl.dev) — Full-stack web intelligence:\n- **scrape** — clean markdown from any URL, JSON schema extraction\n- **search** — web search with snippet scraping\n- **crawl** — multi-page extraction with depth control\n- **map** — discover URLs on a domain\n- **agent** — autonomous research across unknown pages\n- **interact** — click, fill forms, navigate after scrape\n- **monitor** — recurring diffs (markdown or JSON path tracking)\n- **parse** — local PDF/Word/Excel → markdown\n\n> **Use case**: Documentation extraction, competitive research, content monitoring, data pipeline ingestion.\n\n---\n\n### 🐙 GitHub\n```\nrepos  →  issues  →  PRs  →  code review  →  actions  →  releases\n```\n[🐙 MCP GitHub Server](https://github.com/modelcontextprotocol/servers/tree/main/src/github) — Full GitHub lifecycle:\n- Repositories: clone, create, fork, manage remotes\n- Pull requests: diff review, inline comments, merge\n- Issues: create, triage, label, assign, search\n- Actions: workflow status, logs\n- Releases, gists, discussions, projects\n- Multi-toolset architecture: granular permission scoping\n\n> **Use case**: PR code review automation, issue triage, release management.\n\n---\n\n### 🤗 HuggingFace\n```\nhub  →  spaces  →  papers  →  datasets  →  jobs\n```\n[🤗 huggingface.co/mcp](https://huggingface.co/docs/hub/en/mcp) — AI/ML ecosystem:\n- **Hub search** — models, datasets, Spaces by query/filter\n- **Repo details** — overview, dataset structure, preview rows\n- **Space search** — semantic search, MCP-enabled Spaces, dynamic invocation\n- **Paper search** — ML research papers on HF hub\n- **HF Jobs** — remote compute (Python/UV/Docker) for batch inference, evaluation\n- **Doc search** — HF library documentation\n- **Duplicate Space** — copy Spaces (free CPU/GPU tier)\n\n> **Use case**: Model discovery, dataset exploration, batch inference on remote compute.\n\n---\n\n### 🏆 Kaggle\n```\ncompetitions  →  datasets  →  models  →  notebooks  →  leaderboards\n```\n[🏆 kaggle.com/docs/mcp](https://www.kaggle.com/docs/mcp) — Data science & ML competitions:\n- **Competitions** — search, metadata, data files, submissions, leaderboards\n- **Datasets** — search, info, file listing, metadata update, upload\n- **Models** — list, create, update, variations, version files\n- **Notebooks** — create session, save+run, output download, status\n- **Forums** — topics, comments, write-ups, hackathon tracks\n- **Benchmarks** — create tasks, leaderboards\n\n> **Use case**: Competition analysis, dataset ingestion, model versioning, write-up research.\n\n---\n\n### 📊 W&B (Weights & Biases)\n```\nruns  →  artifacts  →  traces  →  reports  →  registry\n```\n[📊 github.com/wandb/wandb-mcp-server](https://github.com/wandb/wandb-mcp-server) — ML experiment & LLM observability:\n- **Runs**: query, filter, sort, history sampling, diagnosis\n- **Artifacts**: list versions, compare diffs, lineage\n- **Traces** (Weave): LLM call tracing, eval summary, schema inference\n- **Reports**: create markdown reports with native chart panels\n- **Registry**: list entities, registries, collections, versions\n- **Analysis**: log computed data, diagnose run health\n\n> **Use case**: Experiment tracking, LLM observability, eval dashboards, model registry.\n\n---\n\n### 📄 arXiv\n```\nsearch  →  abstract  →  download  →  read  →  citation graph  →  alerts\n```\n[📄 github.com/blazickjp/arxiv-mcp-server](https://github.com/blazickjp/arxiv-mcp-server) — Academic research:\n- **Search** — advanced query syntax, field filters, category scoping\n- **Abstract** — quick relevance check before download\n- **Download** → **Read** — full paper in markdown\n- **Citation graph** — Semantic Scholar references/citations\n- **Semantic search** — similarity search across downloaded library\n- **Watch topics** — standing alerts for new papers\n\n> **Use case**: Literature review, citation tracking, research monitoring.\n\n---\n\n### 🌐 Browser Use\n```\nnavigate  →  click  →  type  →  extract  →  screenshot  →  scroll\n```\n[🌐 github.com/browser-use/browser-use](https://github.com/browser-use/browser-use) — Headless browser with full interaction:\n- Multi-tab management (list, switch, close tabs)\n- Element interaction via index or coordinates\n- HTML extraction, content query, screenshots\n- Session management (list, close sessions)\n- Agent mode for complex multi-step tasks\n\n> **Use case**: Dynamic SPA content, login-required pages, form automation.\n\n---\n\n### 📁 Filesystem\n```\nread  →  write  →  search  →  tree  →  edit  →  move\n```\n[📁 MCP Filesystem Server](https://github.com/modelcontextprotocol/servers/tree/main/src/filesystem) — Local file operations:\n- Read/write files, directory listing with sizes\n- Recursive tree view, file search with glob patterns\n- Line-based edit with diff output, file move/rename\n- Media file read (images, audio as base64)\n\n> **Use case**: Code analysis, file manipulation, project exploration.\n\n---\n\n### 📚 Context7\n```\nresolve library  →  query docs\n```\n[📚 mcp.context7.com](https://mcp.context7.com) — Up-to-date library documentation:\n- **Resolve Library ID** — map package name to Context7 ID\n- **Query Docs** — retrieve current docs + code examples with versioning\n\n> **Use case**: Framework documentation queries (Next.js, React, Tailwind, etc.).\n\n---\n\n### 🔗 Git\n```\nstatus  →  diff  →  log  →  branch  →  commit  →  add\n```\n[🔗 mcp-server-git](https://github.com/modelcontextprotocol/servers/tree/main/src/git) — Local git operations:\n- Status, staged/unstaged diffs, commit logs\n- Branch creation, checkout, listing\n- Commit with message, add files, reset staged\n- Diff between branches/commits\n\n> **Use case**: Local version control within agent workflow.\n\n---\n\n### 🕐 Time\n```\ncurrent time  →  convert time\n```\n[🕐 MCP Time Server](https://github.com/modelcontextprotocol/servers/tree/main/src/time) — Timezone utilities:\n- Get current time in any IANA timezone\n- Convert between timezones\n\n---\n\n### 🧠 Sequential Thinking\n```\nstructured reasoning\n```\n[🧠 @modelcontextprotocol/server-sequential-thinking](https://github.com/modelcontextprotocol/servers/tree/main/src/sequentialthinking) — Meta-cognitive reasoning tool:\n- Dynamic, reflective problem-solving\n- Branch, revise, backtrack through thoughts\n- Hypothesis generation and verification\n\n> **Use case**: Complex problem decomposition, debugging, planning.\n\n---\n\n### ⚡ Supabase\n```\ndatabase  →  auth  →  storage  →  realtime\n```\n[⚡ supabase.com/docs/guides/platform/mcp](https://supabase.com/docs/guides/platform/mcp) — Backend-as-a-service via MCP:\n- PostgreSQL queries, table management\n- Authentication, user management\n- File storage\n- Realtime subscriptions\n\n---\n\n## 📊 By the Numbers\n\n```\nTotal MCP Servers  :  13\nTotal Tools        : ~180\nPrimary Agent      :  DeepSeek V4 Flash\nHost               :  macOS (Apple Silicon)\nCLI                :  Hermes Agent → OpenCode Go\n```\n\n---\n\n## 🧠 Why This Stack\n\nThis isn't random — each server fills a role:\n\n```\nResearch    :  arXiv + HuggingFace Papers + Firecrawl Search\nDevelopment :  GitHub + Git + Filesystem + Context7\nData/ML     :  Kaggle + HuggingFace Hub + W&B\nAutomation  :  Browser Use + Firecrawl Agent + HF Jobs\nReasoning   :  Sequential Thinking + Time\nBackend     :  Supabase\n```\n\n---\n\n> **Tip**: Stack evolves. This is a snapshot; new servers get added as workflows demand.\n>\n> _Part of my dev workflow — Enes Demir_\n", "url": "https://wpnews.pro/news/mcp-tool-stack-13-servers-180-tools-ai-native-dev-workflow", "canonical_source": "https://gist.github.com/EnesDemir143/e41e5b6df8f2523c1240514132feb51f", "published_at": "2026-05-22 09:40:47+00:00", "updated_at": "2026-05-22 09:52:27.843290+00:00", "lang": "en", "topics": ["developer-tools", "artificial-intelligence", "machine-learning", "open-source", "data"], "entities": ["Model Context Protocol", "Firecrawl", "GitHub", "Hugging Face", "Hermes Agent", "DeepSeek V4 Flash"], "alternates": {"html": "https://wpnews.pro/news/mcp-tool-stack-13-servers-180-tools-ai-native-dev-workflow", "markdown": "https://wpnews.pro/news/mcp-tool-stack-13-servers-180-tools-ai-native-dev-workflow.md", "text": "https://wpnews.pro/news/mcp-tool-stack-13-servers-180-tools-ai-native-dev-workflow.txt", "jsonld": "https://wpnews.pro/news/mcp-tool-stack-13-servers-180-tools-ai-native-dev-workflow.jsonld"}}