lemonsqueezy-mcp: Instructions file for Claude Code

CLAUDE.md

lemonsqueezy-mcp CLAUDE.md is an instructions file for Claude Code from YawLabs/lemonsqueezy-mcp. It costs 2,309 tokens per session, scanned A, original, MIT.

A set of instructions for coding agents working on a Lemon Squeezy MCP server. Lemon Squeezy is a service for selling digital products and managing subscriptions, customers, and licenses.

In plain words
What is it for?
Use it when developing or reviewing this server, including its Lemon Squeezy API client, tool modules, guardrails, build, linting, and tests.
Why use it?
It gives the agent the project's structure, build rules, testing commands, and safety checks so changes follow the repository's conventions.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md.

This is YawLabs/lemonsqueezy-mcp's own configuration. It tells Claude Code how to work on lemonsqueezy-mcp 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 lemonsqueezy-mcp configures →

Reuse

Borrowing it

Nothing to install: this file belongs to YawLabs/lemonsqueezy-mcp. 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/YawLabs/lemonsqueezy-mcp/main/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/YawLabs/lemonsqueezy-mcp

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 lemonsqueezy-mcp CLAUDE.md

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

Your own site · 80×15
<a href="https://agentmods.dev/instructions/yawlabs/lemonsqueezy-mcp/claude-md"><img src="https://agentmods.dev/badge/instructions/yawlabs/lemonsqueezy-mcp/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 2,309 This file is loaded in full into every session.
When invoked 2,309 The same file — it is already loaded in full.
Security scan A 1 finding. 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 $0.02309 $0.02309
Opus 5.5 $0.00924 $0.00924
Sonnet 5.5 $0.00462 $0.00462
Haiku 4.5 $0.00231 $0.00231

Measured 15d ago against content hash 2a62a2b17607, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

lemonsqueezy-mcp CLAUDE.md scanned grade A with 1 finding 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 15d 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

Releases run **locally** from a clean checkout of `main` via `./release.sh <version>`, which does steps 1-8 on the workstation -- lint, test, build, bump, commit, tag, push, npm publish, GitHub release, MCP Registry publ
CLAUDE.md · 55 lines

How it starts

The opening of the file, as written. The whole thing — 55 lines — stays where its author put it; the contents beside it link to each section on GitHub.

lemonsqueezy-mcp

LemonSqueezy MCP server — manage your store, subscriptions, customers, and licenses from AI assistants.

Architecture

  • src/index.ts — Entry point. Registers all tools with McpServer, handles version subcommand. Must stay free of top-level await (the oam compile binary build emits CJS).
  • src/api.ts — LemonSqueezy API client. Bearer token auth, JSON:API format, 30s timeout. Also exports licenseRequest for the License API (different auth), plus the shared buildQuery / buildInvoiceQuery / cross-store-note helpers used by the tool modules.
  • src/wrapper.ts — Sits between the SDK's registered handler and each tool handler. Runs the guardrails in a load-bearing order, then logs and audits. createToolHandler is the only production caller path.
  • src/guardrails.ts — Store allowlist, refund cap, destructive + per-authority-class rate limits. All opt-in via env.
  • src/tools/*.ts — Tool definitions as exported arrays. Each file covers one API resource domain.

Build

  • Bundler: esbuild (build.mjs) — single dist/index.js with zero runtime deps
  • Type checking: tsc (separate pass before esbuild)
  • Linter: Biome
  • Tests: Node.js built-in test runner (node --test)
  • TypeScript: Strict mode, ES2022 target, Node16 module resolution

Key patterns

  • Tools are arrays of { name, description, authorityClass, annotations, inputSchema, handler } objects
  • All tool names prefixed with ls_
  • Zod schemas for input validation with .describe() for each field
  • Every tool has MCP annotations: readOnlyHint, destructiveHint, idempotentHint, openWorldHint
  • JSON:API query params built via buildQuery({ include, filter, page })
  • Version injected at build time via esbuild define
  • Optional per-tool hooks on the wrapper: isDestructive(input) (per-call destructive verdict), requiredFilters (store-allowlist scoping for list-by-parent tools), and preflight(input) (input guardrail that must reject before the rate limiters — the refund cap uses it)
  • A list tool with neither a storeId field nor requiredFilters is completely ungated by the store allowlist. tools.test.ts fails on any new one; the two known cases must disclose it via crossStoreUngatedNote()
  • Update tools reject an ID-only PATCH locally rather than paying a round-trip for an upstream 422, throwing ToolInputError so the wrapper logs it as validation_error (client mistake) rather than exception (something faulted) or guardrail_block (operator policy)
  • redact.ts guards cycles with an ancestor path plus a memo of completed subtrees — the ancestor set alone is O(paths), which is exponential on a shared-reference DAG
  • redact.ts masks license keys: a key matching the anchored LICENSE_KEY_RE (licenseKey / license_key / license-key, any case — never licenseKeyId, license_key_id or licenseKeyInstanceId) becomes [REDACTED:last4=XXXX] when its value is a string of 16+ characters, and [REDACTED] otherwise (fails closed). It is the single choke point for audited inputs in every sink (stderr, the audit ring, the lemonsqueezy://audit-log resource) and for api.ts's unrecognized-error-body fallback. The error field never passes through it: api.ts only accepts a string upstream message (extractErrorMessage) and redacts the body otherwise, so keep that string check if you touch the error path. Tests derive the expected mask from the exported maskLicenseKey, not a hard-coded string. A tool that takes a raw license key must never become destructive (audited) without this mask
  • The four predicate tools (ls_update_customer, ls_update_subscription, ls_update_license_key, ls_update_webhook) declare static destructiveHint: true: MCP clients see only the static annotation, and MCP defines false as "only additive updates". isDestructiveCall never reads the hint when a predicate exists, so the server-side destructive limiter and audit log stay per call. tools.test.ts pins the destructiveHint: true set (13), the predicate set (4), the readOnlyHint: false set and the full 64-tool class map as SEMVER.md surface — changing one is a SEMVER decision plus a README update, not a test fixup

Read the full file on GitHub · 55 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. 15d ago Changed · +9 tokens per session 2a62a2b17607
  2. 23d ago Changed · +2 tokens per session f8a2813e7fd9
  3. 1mo ago Changed · +2 lines · +521 tokens per session d8b6d0fb461b
  4. 1mo ago Changed · +272 tokens per session 70ae214e49c3
  5. 2mo ago First seen · 53 lines · 1,505 tokens per session scan A abe1ae3a1705

Subscribe to this mod's changes

lemonsqueezy-mcp CLAUDE.md is an instructions file published in the GitHub repository YawLabs/lemonsqueezy-mcp (3 stars, last pushed 16d ago), licensed MIT. It adds 2,309 tokens to every session, about $0.0092 per session on Opus 5.5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other instructions, from other repositories

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 9,715 tokens

next.js AGENTS.md

AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,461 tokens

vscode oss-third-party-notices.instructions.md

Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).

microsoft/vscode · 5,262 tokens

langchain AGENTS.md

AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.

langchain-ai/langchain · 4,680 tokens

vscode copilot-instructions.md

Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).

microsoft/vscode · 2,750 tokens

deepseek-harness AGENTS.md

AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.

deepseek-ai/deepseek-harness · 4,261 tokens