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 agentmods add skills/laurigates/claude-plugins/mcp-server-authoringnpx skills add laurigates/claude-plugins --skill mcp-server-authoringgit clone --depth 1 https://github.com/laurigates/claude-pluginsWrote 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/laurigates/claude-plugins/mcp-server-authoring)<a href="https://agentmods.dev/skills/laurigates/claude-plugins/mcp-server-authoring"><img src="https://agentmods.dev/badge/skills/laurigates/claude-plugins/mcp-server-authoring.svg" alt="Measured on agentmods" 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 | $0.00050 | $0.03009 |
| Opus 5 | $0.00025 | $0.01504 |
| Sonnet 5 | $0.00010 | $0.00602 |
| Haiku 4.5 | $0.00005 | $0.00301 |
Grade A, and why
mcp-server-authoring 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 6d 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 — 260 lines — stays where its author put it; the contents beside it link to each section on GitHub.
MCP Server Authoring
Producer-side patterns for building a Python Model Context Protocol server
with the official SDK's FastMCP (or the standalone fastmcp package). Applies
to any Python MCP server you own — the guidance is portfolio-independent.
For consuming / installing servers into a project, this is the wrong skill — see the table below.
When to Use This Skill
| Use this skill when... | Use a different skill when... |
|---|---|
| Building or scaffolding a new MCP server | Installing an existing server into .mcp.json → agent-patterns-plugin:mcp-management |
| Adding a tool / resource / prompt to a server | Running MCP compliance checks on a project → configure-plugin:configure-mcp |
| Wiring a server's tests, lint, release-please | An agent calls 50+ MCP tools and needs the code-execution pattern → agent-patterns-plugin:mcp-code-execution |
| Choosing transport (stdio vs HTTP) for a server you own | Managing OAuth for a remote server you consume → mcp-management |
The Build Path
Build a server in this order. Each step's why is the load-bearing part — it is what lets you generalize past the example.
Step 1 — Scaffold the project
uv init --package my-server && cd my-server
uv add "mcp[cli]" # bundled SDK; or `uv add fastmcp` for the standalone package
uv add --group dev pytest ruff ty
Why uv init --package: it lays down a src/ package + pyproject.toml
with a console-script entrypoint, so the server installs as a real command
(uvx my-server) rather than a loose script. That is what a consumer's
.mcp.json will invoke.
Why one dependency choice up front: FastMCP ships two ways — bundled in the
official mcp SDK (from mcp.server.fastmcp import FastMCP) and as the
standalone fastmcp package (from fastmcp import FastMCP). They expose the
same high-level decorator API; pick one and pin it in pyproject.toml so the
import path is stable. Prefer the bundled SDK unless you need a standalone-only
feature.
What ships with it
1 file 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.
- 6d ago First seen · 260 lines · 50 tokens per session scan A 09d0b4963cc9
mcp-server-authoring is a skill published in the GitHub repository laurigates/claude-plugins (58 stars, last pushed today), licensed MIT. It adds 50 tokens to every session and 3,009 once invoked, about $0.0003 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 2026-08-30.
Other skills, from other repositories
intern-symbols-identity
For fast comparison: ensure only one instance of each symbol, use 'is' instead of '==', symbol tables for interpreters.
iterate-with-itertools
For combinatorial iteration: permutations, combinations, cartesian products, without storing all results in memory.
overload-operators-dsl
For domain-specific languages: operator overloading, make Python look like math/domain notation, expression builders.
parse-extract-input
For text processing: extract numbers, words, structured data from messy text using regex patterns, parsing utilities.
pass-function-as-arg
For generic algorithms: strategy pattern, callbacks, configurable behavior. Pass functions as parameters to customize algorithm behavior.
tokenize-then-parse
For interpreters: separate lexing from parsing, token stream to AST, structured text processing in stages.