mcp-loop-build-demo: Instructions file for Claude Code

CLAUDE.md

mcp-loop-build-demo CLAUDE.md is an instructions file for Claude Code from EezyMAcc/mcp-loop-build-demo. It costs 1,807 tokens per session, scanned A, original, MIT.

Repository instructions for building an MCP server, a program that lets an AI client use selected tools, to browse and organize Instagram posts.

In plain words
What is it for?
Use them before changing this repository, especially when adding features, updating tests, or modifying the MCP server and its documentation.
Why use it?
They define the project structure, testing rules, and development workflow so changes do not weaken checks or mix core application logic with the tool interface.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md. Also seen: mentions CLAUDE.md.

This is EezyMAcc/mcp-loop-build-demo's own configuration. It tells Claude Code how to work on mcp-loop-build-demo itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything mcp-loop-build-demo configures →

Reuse

Borrowing it

Nothing to install: this file belongs to EezyMAcc/mcp-loop-build-demo. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/EezyMAcc/mcp-loop-build-demo/main/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/EezyMAcc/mcp-loop-build-demo

Made for: Claude Code.

Wrote 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.

agentmods badge for mcp-loop-build-demo CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/eezymacc/mcp-loop-build-demo/claude-md/github.svg)](https://agentmods.dev/instructions/eezymacc/mcp-loop-build-demo/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/eezymacc/mcp-loop-build-demo/claude-md"><img src="https://agentmods.dev/badge/instructions/eezymacc/mcp-loop-build-demo/claude-md/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.

agentmods 80×15 button for mcp-loop-build-demo CLAUDE.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/eezymacc/mcp-loop-build-demo/claude-md"><img src="https://agentmods.dev/badge/instructions/eezymacc/mcp-loop-build-demo/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 1,807 This file is loaded in full into every session.
When invoked 1,807 The same file — it is already loaded in full.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.01807 $0.01807
Opus 5.5 $0.00723 $0.00723
Sonnet 5.5 $0.00361 $0.00361
Haiku 4.5 $0.00181 $0.00181

Measured 2mo ago against content hash 4fa98e7ded9f, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

mcp-loop-build-demo CLAUDE.md 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 2mo 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.

CLAUDE.md · 183 lines

How it starts

The opening of the file, as written. The whole thing — 183 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

Operating rules for autonomous work on this repository. Read this fully before touching any code. These rules are non-negotiable and override any instinct to "just make the gate pass."


1. What this project is

A Model Context Protocol (MCP) server that lets an LLM client (e.g. Claude Desktop) browse and curate posts from a hand-picked list of Instagram tattoo artists, via Instagram's Business Discovery API.

Architecture is a clean two-layer split:

  • core — all real logic: domain models, typed errors, repositories (JSON-file persistence), the Graph API client, image processing, and the services that orchestrate them. Knows nothing about MCP.
  • server — a thin FastMCP adapter that exposes core as MCP tools. Holds no business logic.

This split exists so a phase-2 GUI can later reuse core directly without a rewrite. Do not leak MCP concepts into core, and do not put logic in server.

Full build plan and tool contracts live in PLAN.md. This file governs how you work; PLAN.md governs what you build.


2. Golden rules (anti-cheat — violating any of these fails the run)

  1. Never edit, delete, skip, weaken, or xfail a test to make a gate pass. If a test is wrong, stop and record it in BLOCKERS.md — do not "fix" it by making it assert less.
  2. Never weaken tooling config to pass a gate. ruff, mypy --strict, and the coverage floor (--cov-fail-under=90) are fixed. Do not edit pyproject.toml lint/type/coverage settings to get green.
  3. Never make a live network call in the test path. All HTTP to Instagram is mocked with respx. The one real integration test is gated behind RUN_LIVE=1 and must never be run by you.
  4. Never commit secrets. No tokens, no IDs, no .env. Only .env.example with placeholder values is committed.
  5. Never weaken coverage by adding blanket # pragma: no cover. The only permitted pragma is on the stdio entrypoint main() in server/app.py.

Read the full file on GitHub · 183 lines

Changes

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.

  1. 2mo ago First seen · 183 lines · 1,807 tokens per session scan A 4fa98e7ded9f

Subscribe to this mod's changes

mcp-loop-build-demo CLAUDE.md is an instructions file published in the GitHub repository EezyMAcc/mcp-loop-build-demo (0 stars, last pushed 5mo ago), licensed MIT. It adds 1,807 tokens to every session, about $0.0072 per session on Opus 5.5. 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.

Related

mcp-loop-build-demo CLAUDE.md compared with instructions from other repositories

mcp-loop-build-demo CLAUDE.md next to 6 instructions from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
mcp-loop-build-demo CLAUDE.mdEezyMAcc/mcp-loop-build-demoInstructions file05mo agoMITA1,807Claude Code
vscode buildNext.instructions.mdmicrosoft/vscodeInstructions file196ktodayMITA9,719GitHub Copilot
next.js AGENTS.mdvercel/next.jsInstructions file144ktodayMITA8,130Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscodeInstructions file196ktodayMITA5,685GitHub Copilot
langchain AGENTS.mdlangchain-ai/langchainInstructions file150ktodayMITA4,680Codex, OpenCode
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harnessInstructions file283ktodayMITA4,342Codex, OpenCode
vscode copilot-instructions.mdmicrosoft/vscodeInstructions file196ktodayMITA3,153GitHub Copilot