Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add cyberreinxy/mcp-server-skill --skill building-mcp-serversgit clone --depth 1 https://github.com/cyberreinxy/mcp-server-skillWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/cyberreinxy/mcp-server-skill/building-mcp-servers)<a href="https://agentmods.dev/skills/cyberreinxy/mcp-server-skill/building-mcp-servers"><img src="https://agentmods.dev/badge/skills/cyberreinxy/mcp-server-skill/building-mcp-servers/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/cyberreinxy/mcp-server-skill/building-mcp-servers"><img src="https://agentmods.dev/badge/skills/cyberreinxy/mcp-server-skill/building-mcp-servers.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00160 | $0.02255 |
| Opus 5 | $0.00080 | $0.01128 |
| Sonnet 5 | $0.00032 | $0.00451 |
| Haiku 4.5 | $0.00016 | $0.00226 |
Grade A, and why
building-mcp-servers scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 107 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Building MCP Servers
A practical, security-first guide to building Model Context Protocol (MCP) servers — servers that expose tools, resources, and prompts to AI agents like Claude. Written with a database-first lens (Postgres especially), because that's where most of the sharp edges live, but the architecture section applies to any MCP server.
This skill exists because most "MCP server tutorials" stop at npx create-mcp-server and a toy add(a, b) tool. Real MCP servers — the kind that touch a production database — fail in specific, recurring ways: unparameterized SQL, stale schema caches, install scripts that aren't pinned, stack traces leaked to the client, and scope creep in PRs that bundle an unrequested dashboard into a "security fix." This skill encodes the fixes for all of those.
Step 0: Figure out what's actually being asked
Before writing code, classify the task — it changes what you read next:
| Task | Go to |
|---|---|
| Building a new MCP server from scratch | Sections 1–4, then the relevant reference file |
| Adding a tool/resource to an existing server | Section 2 |
| Server touches a SQL database | references/postgres-database-servers.md (do this even if the user didn't ask about security — read it first) |
| Reviewing a PR or contributor patch to an MCP server | references/security-review-checklist.md |
| Testing, packaging, or getting the server into Claude Desktop/Claude Code | references/testing-and-distribution.md |
| Something's failing silently or the client can't see the server | references/debugging.md |
Don't skip the security reference just because the user only asked for a feature. Database MCP servers are agent-facing and internet-facing in ways normal backend code isn't — an LLM will construct arbitrary queries against your tools, including adversarial ones from prompt-injected content it read elsewhere.
1. Core protocol model
MCP servers expose three primitive types over a JSON-RPC transport. Get this taxonomy right before writing any code — most bad MCP servers are bad because they use the wrong primitive.
What ships with it
4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 107 lines · 160 tokens per session scan A 79e9c130aa0e
building-mcp-servers is a skill published in the GitHub repository cyberreinxy/mcp-server-skill (1 stars, last pushed 1mo ago), licensed MIT. It adds 160 tokens to every session and 2,255 once invoked, about $0.0008 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
api-security
Use for authorized security assessment of REST, GraphQL, WebSocket, or SOAP APIs, including discovery, authentication, authorization, rate-limit, and CI/CD testing.
protocol-reverse
Use for authorized reverse engineering of custom binary protocols, Protobuf/gRPC, WebSocket frames, and PCAP-driven protocol recovery.
competition-graphql-rpc-drift
Internal downstream skill for ctf-sandbox-orchestrator. CTF-sandbox workflow for GraphQL schemas, persisted queries, RPC manifests, generated clients, OpenAPI drift, hidden operations, and contract-to-handler mismatches. Use when the user asks to inspect GraphQL or RPC requests, compare client contracts to live…
competition-queue-worker-drift
Internal downstream skill for ctf-sandbox-orchestrator. CTF-sandbox workflow for queues, async workers, cron jobs, delayed tasks, retry behavior, worker-only config drift, and payload-to-side-effect chains. Use when the user asks to trace a queue payload, inspect async job execution, explain worker-only behavior…
competition-template-render-path
Internal downstream skill for ctf-sandbox-orchestrator. CTF-sandbox workflow for SSR, template rendering, route loaders, hydration payloads, server-client render boundaries, and template-to-handler enforcement gaps. Use when the user asks to inspect SSR or template routes, trace render context or hydration data…
competition-websocket-runtime
Internal downstream skill for ctf-sandbox-orchestrator. CTF-sandbox workflow for WebSocket and SSE handshakes, auth material, subscription state, realtime message schemas, reconnect behavior, and frame-driven runtime effects. Use when the user asks to inspect a WebSocket or SSE handshake, decode frames, trace…