Cockpit Foundation
- 10 pillar surfaces
- 5 explainers · 4 FAQ · 1 recipe
- MCP cron pipeline
- Cmd+K palette
- Stack diagram + coverage matrix
AGENTS // FIELD GUIDE // ONLINE cron 4h ago
A field guide for building with agents.
Vendor docs are pitch. Hacker News is a lottery. Honest MCP reviews, side-by-side vendor takes, and recipes with measured numbers — by an engineer who deploys this for a living.
- 179k★
- tracked
- 13
- mcps reviewed
- 1
- recipes
- 2
- vendors
- 4
- standards
- 14
- safety lenses
ECOSYSTEM // 24H SNAPSHOT
13 tracked mcps 179k★ total stars +204 stars / 7d 10 alive 1 slowing 34d avg freshness Productivity top category 8 categories
Cron-fed updates from the agentic ecosystem.
Vendor releases, protocol changes, MCP star deltas, recently committed servers — auto-refreshed every 6 hours.
- Anthropic ships Computer Use v3 with 30% lower latency
- MCP Spec 2.0 RFC drops — auth + streaming improvements
- Claude Skills hits GA, marketplace open
- AutoGen in maintenance mode — successor: Microsoft Agent Framework
- Microsoft Agent 365 GA next quarter
Everything we cover, with progress visible.
Each pillar shows current/target. Some are complete (Standards 4/4), others growing (MCPs 13/30).
Automate YouTube SEO across many videos
Three MCPs chained: rewrite titles in Sush's voice, push updates, verify cache. Scaffolded — measurements pending Sush's verified runs.
Several hours / month manual Single-digit minutes · Roughly low single-digit cents per video (token usage)
Open recipe → /recipes/youtube-seo-automation/ github-mcp-server
Official GitHub MCP server. Browse, search, and modify repos, issues, PRs, Actions from any MCP-aware agent. Maintained by GitHub.
Read review → /mcp/github-mcp/The core surfaces.
LIVE
MCP Health Board · top 8
Server Category Stars Δ 7d Trend Last commit Status
github-mcp-server Dev 29.8k +74 <1d ago Alive playwright-mcp Browser 32.5k +70 1d ago Alive filesystem-mcp in modelcontextprotocol/servers / src/filesystem Dev
(85.6k repo)
— — 1d ago Alive postgres-mcp in modelcontextprotocol/servers / src/postgres Data
(85.6k repo)
— — 1d ago Alive slack-mcp in modelcontextprotocol/servers / src/slack Productivity
(85.6k repo)
— — 1d ago Alive cloudflare-mcp Cloud 3.7k +6 1w ago Alive notion-mcp Productivity 4.3k +8 1mo ago Slowing figma-mcp Design 14.7k +24 2w ago Alive HONEST FAQ
Big questions
Q.01 Is agentic AI hype? Some of it — most consumer demos are. The honest test: does it execute multi-step work without you holding its hand? Q.02 Will MCP survive 18 months? Probably yes. Claude Desktop adoption, GitHub MCP at 29.5k★, and OpenAI shipping a compatible Apps SDK adapter — that's a moat. Q.03 What can I build with $0? A capable research + writing + scheduling stack — Claude Free + GitHub MCP + Brave Search MCP + Filesystem MCP. Real workflows, no credit card. Q.04 Why do agents still hallucinate? They're trained to predict plausible next-tokens, not to verify truth. RAG and tool-calls reduce it; nothing eliminates it. Always verify.
VENDOR FILES
Side-by-side, no spin
Catalog + Reviews
- 5 measured recipes
- 30 MCP reviews
- Vendor expansion (OpenAI + Google)
- Search index (Pagefind)
- Public health badges
Standalone Property
- Dedicated domain (agentic.dev or similar)
- Templates pillar with public repos
- Career pillar
- Lab tools (cost calc, comparator)
- 30+ Excalidraw diagrams
QUICK INSTALL
Add an MCP in 30 seconds
GitHub Dev
npx -y @github/mcp-server Playwright Browser
npx -y @playwright/mcp Filesystem Dev
npx -y @modelcontextprotocol/server-filesystem Postgres Data
npx -y @modelcontextprotocol/server-postgres CONFIG → ~/.config/Claude/claude_desktop_config.json
$ npx -y @github/mcp-server [INSTALL]
$ cron run --refresh-mcp-stars [CRON]
$ git push origin main [DEPLOY]
$ agentic stack diagram --layer LLM [INSPECT]
$ curl agents.aguidetocloud.com/api/health [PING]
$ recipe run --id youtube-seo-automation [RUN]
$
EXPLAINERS
Plain English, no jargon
L01 Landscape 1 / 4 25% EARLY R02 Recipes 1 / 5 20% EARLY M03 MCP Servers 13 / 30 43% EARLY T04 Tools 0 / 12 0% EARLY V05 Vendors 2 / 4 50% IN-PROGRESS S06 Safety 14 / 14 100% COMPLETE P07 Standards 4 / 4 100% COMPLETE L08 Learn 5 / 10 50% IN-PROGRESS X09 Lab 1 / 6 17% EARLY Q10 FAQ 4 / 10 40% EARLY