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 instructions/wbopan/things-mcp/claude-mdgit clone --depth 1 https://github.com/wbopan/things-mcpWrote 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/instructions/wbopan/things-mcp/claude-md)<a href="https://agentmods.dev/instructions/wbopan/things-mcp/claude-md"><img src="https://agentmods.dev/badge/instructions/wbopan/things-mcp/claude-md.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 | $0.01387 | $0.01387 |
| Opus 5 | $0.00694 | $0.00694 |
| Sonnet 5 | $0.00277 | $0.00277 |
| Haiku 4.5 | $0.00139 | $0.00139 |
Grade A, and why
things-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 5d 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.
Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
- Uses `child_process.exec` with `open` command for URL schemes How it starts
The opening of the file, as written. The whole thing — 132 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
Development Commands
Essential Commands
npm run build # Compile TypeScript to build/ directory
npm test # Run Jest tests with ES module support
npm run dev # Watch mode for development
npm start # Run the MCP server
npm run inspector # Launch MCP Inspector for testing tools
Testing Commands
npm test -- --testNamePattern="URL Builder" # Run specific test suite
npm test -- tests/server.test.ts # Run specific test file
npm run test:watch # Watch mode for tests
Memories
- Do not run npm run inspector on your own
- github repo is https://github.com/BMPixel/things-mcp
Architecture Overview
MCP Server Design
This is a Model Context Protocol (MCP) server that integrates with Things.app on macOS via URL scheme automation. The architecture follows these key patterns:
Core Flow:
- MCP tools receive structured parameters (validated with Zod schemas)
- Tools convert parameters to Things URL scheme format via
url-builder.ts - URLs are executed using macOS
opencommand to trigger Things.app actions - All operations return MCP-compliant content responses
Tool Registration Pattern:
Each tool follows a consistent pattern in src/tools/:
- Export a
registerXxxTool(server: McpServer)function - Use Zod schemas for input validation with
.shapeproperty - Call
server.tool(name, schema.shape, handler)to register - Return
{ content: [{ type: "text", text: "..." }] }format
Key Components
URL Construction (src/utils/url-builder.ts):
buildThingsUrl(): Converts parameters to Things URL schemeopenThingsUrl(): Executes URLs via macOSopencommand (async/await)- Uses
encodeURIComponent()for proper percent encoding (spaces become%20)
Authentication (src/utils/auth.ts):
- Update operations require
THINGS_AUTH_TOKENenvironment variable - Use
requireAuthToken()for operations that modify existing items - Token obtained from Things.app Preferences → General → Enable Things URLs
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.
- 5d ago First seen · 132 lines · 1,387 tokens per session scan A 142aad51b612
things-mcp CLAUDE.md is an instructions file published in the GitHub repository wbopan/things-mcp (3 stars, last pushed 8mo ago), licensed MIT. It adds 1,387 tokens to every session, about $0.0069 per session on Opus 5. A static security scan graded it A with 1 finding (runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
things-cloud-mcp AGENTS.md
Instructions for wbopan/things-cloud-mcp, covering agents.md, build and test, architecture and safety invariants, mcp tools and wire format.
things-cloud-mcp CLAUDE.md
Instructions for wbopan/things-cloud-mcp, covering claude.md, build & run, architecture, mcp tool design and coding patterns.
mcp-server-things CLAUDE.md
Claude Code instructions for ebowman/mcp-server-things, covering things 3 mcp server - ai assistant instructions, project overview, ✨ latest features (v1.7.0), architecture and development guidelines.
mcp-server-things AGENTS.md
AGENTS.md instructions for ebowman/mcp-server-things, covering agent instructions, quick reference, non-interactive shell commands, force overwrite without prompting and for recursive operations.
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
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).