What is mcp-anything CLAUDE.md?

Project instructions for MCP-Anything, a system that turns a plain-language description and a data source such as an API description or database schema into an MCP server and related files.

Use them when developing or running MCP-Anything, including its command-line tool, code generators, skill bundle creation, and validation tests.

They give an agent the project’s purpose, development commands, architecture, and main generation stages in one place.

How gabrielekarra/mcp-anything configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to gabrielekarra/mcp-anything. 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/gabrielekarra/mcp-anything/main/CLAUDE.md

Clone the repo

git clone --depth 1 https://github.com/gabrielekarra/mcp-anything

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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/gabrielekarra/mcp-anything/claude-md/github.svg)](https://agentmods.dev/instructions/gabrielekarra/mcp-anything/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/gabrielekarra/mcp-anything/claude-md"><img src="https://agentmods.dev/badge/instructions/gabrielekarra/mcp-anything/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.

Your own site · 80×15

<a href="https://agentmods.dev/instructions/gabrielekarra/mcp-anything/claude-md"><img src="https://agentmods.dev/badge/instructions/gabrielekarra/mcp-anything/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 5,812 This file is loaded in full into every session.

When invoked 5,812 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.

Model	Per session	Once invoked
Fable 5.1	$0.05812	$0.05812
Opus 5	$0.02906	$0.02906
Sonnet 5	$0.01162	$0.01162
Haiku 4.5	$0.00581	$0.00581

Measured 9d ago against content hash d6fdaffa7d15, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

mcp-anything 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 9d 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 · 374 lines

How it starts

The opening of the file, as written. The whole thing — 374 lines — stays where its author put it; the contents beside it link to each section on GitHub.

MCP-Anything

What this project does

LLM-driven pipeline that takes a customer's domain brief (use-cases in natural language) and a data source (OpenAPI / gRPC / DB schema / SDK), and produces a fully-implemented, optimized MCP server plus skill bundle and validation artifacts.

Three output backends: Python/FastMCP, TypeScript/mcp-use, and TypeScript/Skybridge (MCP + ChatGPT App with React views).

Legacy path: mcp-anything generate <path> (codebase scanner, backwards-compatible). Domain path: mcp-anything build --brief <brief.yaml> (new, recommended).

Development

Install: pip install -e ".[dev,llm]"
Run tests: pytest tests/ -v
Run CLI: mcp-anything --help

Architecture

Domain Pipeline (new, primary)

5 phases: DOMAIN_MODELING → TOOL_DESIGN → EMIT → SKILL_BUNDLE → VALIDATION_HARNESS

Phase 1 (domain_modeling.py): LLM reads brief + data source → domain_model.json
Phase 2 (tool_design.py): LLM shapes tools per 2026 rules → tool_spec.yaml
Phase 3 (emit/python_fastmcp/, emit/typescript_mcp_use/, or emit/typescript_skybridge/): code generation
Phase 4 (skill_bundle.py): LLM generates SKILL.md + quick_queries.json
Phase 5 (validation_harness.py): LLM generates eval_cases.json, optional live eval
Output contract: CONTRACT.md — 29 testable items (C-01..C-29) both emitters must satisfy
Conformance suite: src/mcp_anything/conformance/ with parity assertion and CI reporter

Legacy Pipeline (preserved, unchanged)

5 phases: ANALYZE → DESIGN → IMPLEMENT → DOCUMENT → PACKAGE

Static detectors + optional LLM analysis
Jinja2 templates in src/mcp_anything/codegen/templates/
mcp-anything generate <path> — unchanged

2026 Stack Defaults (wired into domain pipeline)

Group CRUD: ≥3 CRUD ops on same resource → single manage_X(operation=...) tool
Composed tools: multi-step workflows promoted to single callable
Progressive disclosure: disclosure_level on every tool, verbose-only tools hidden by default
Compact responses: every tool has verbose flag (C-10)
Discovery endpoint: GET /.well-known/mcp (C-01..C-03)
Telemetry: anonymized per-call logging, MCP_TELEMETRY_ENDPOINT env var (C-19..C-20)
Docker: Dockerfile in every generated server (C-17)
SKILL.md: required sections Overview, Tools, Usage Patterns, Gotchas, Recipes, Anti-patterns (C-22)

Read the full file on GitHub · 374 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.

9d ago First seen · 374 lines · 5,812 tokens per session scan A d6fdaffa7d15

Subscribe to this mod's changes

mcp-anything CLAUDE.md is an instructions file published in the GitHub repository gabrielekarra/mcp-anything (43 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 5,812 tokens to every session, about $0.0291 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.

Building agents Code Generation ai anthropic automation claude cli

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 · 7,296 tokens

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.

openai/codex · 5,153 tokens

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 · 6,785 tokens

mcp-anything: Instructions file for Claude Code