smart-ralph CLAUDE.md

smart-ralph CLAUDE.md is an instructions file for coding agents from tzachbon/smart-ralph. It costs 1,907 tokens per session, scanned A, original, MIT.

Repository instructions for Claude Code covering safety rules, simple coding, and focused changes. A pull request is a proposed set of code changes for review before merging.

In plain words
What is it for?
Use them when planning changes, deciding whether to ask questions, editing code, and handling pull requests, branches, or merges.
Why use it?
They reduce accidental destructive actions and discourage unnecessary features, abstractions, and unrelated edits.

Instructions file

Install

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.

agentmods
npx agentmods add instructions/tzachbon/smart-ralph/claude-md
Clone the repo
git clone --depth 1 https://github.com/tzachbon/smart-ralph

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 smart-ralph CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/tzachbon/smart-ralph/claude-md.svg)](https://agentmods.dev/instructions/tzachbon/smart-ralph/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/tzachbon/smart-ralph/claude-md"><img src="https://agentmods.dev/badge/instructions/tzachbon/smart-ralph/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 1,907 This file is loaded in full into every session.
When invoked 1,907 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
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 $0.01907 $0.01907
Opus 5 $0.00954 $0.00954
Sonnet 5 $0.00381 $0.00381
Haiku 4.5 $0.00191 $0.00191

Measured 6d ago against content hash c3b392300534, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

smart-ralph 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 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.

CLAUDE.md · 182 lines

How it starts

The opening of the file, as written. The whole thing — 182 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.

⛔ CRITICAL SAFETY RULES

  1. NEVER merge PRs without explicit user permission — If unsure whether to merge, the answer is NO
  2. NEVER close PRs without explicit user permission — Only fix conflicts, push changes, create PRs
  3. NEVER delete branches on remote without explicit user permission
  4. Ask before any destructive action — When in doubt, ask the user

Karpathy Coding Rules

Four rules for all agents and code generation. Non-negotiable.

1. Think Before Coding

  • State assumptions explicitly. If uncertain, ask.
  • Multiple interpretations? Present them, don't pick silently.
  • Simpler approach exists? Say so. Push back when warranted.
  • Something unclear? Stop. Name what's confusing. Ask.

2. Simplicity First

  • No features beyond what was asked.
  • No abstractions for single-use code.
  • No speculative "flexibility" or "configurability".
  • 200 lines that could be 50? Rewrite.
  • Test: "Would a senior engineer say this is overcomplicated?"

3. Surgical Changes

  • Don't "improve" adjacent code, comments, or formatting.
  • Don't refactor what isn't broken.
  • Match existing style, even if you'd do it differently.
  • Remove only dead code YOUR changes created.
  • Every changed line must trace to the user's request.

4. Goal-Driven Execution

  • "Add validation" -> Write tests for invalid inputs, make them pass.
  • "Fix the bug" -> Write reproducing test, make it pass.
  • "Refactor X" -> Ensure tests pass before and after.
  • Define success criteria. Loop until verified.

Overview

Smart Ralph is a Claude Code plugin for spec-driven development. It transforms feature requests into structured specs (research, requirements, design, tasks) then executes them task-by-task with fresh context per task.

Development

# Test plugin locally
claude --plugin-dir ./plugins/ralph-specum

# Test the workflow
/ralph-specum:start test-feature Some test goal

Read the full file on GitHub · 182 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. 6d ago First seen · 182 lines · 1,907 tokens per session scan A c3b392300534

Subscribe to this mod's changes

smart-ralph CLAUDE.md is an instructions file published in the GitHub repository tzachbon/smart-ralph (533 stars, last pushed 2d ago), licensed MIT. It adds 1,907 tokens to every session, about $0.0095 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.