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/thecodingend/rails-starter/claude-mdgit clone --depth 1 https://github.com/thecodingend/rails-starterWhat 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 | $0.01060 | $0.01060 |
| Opus 5 | $0.00530 | $0.00530 |
| Sonnet 5 | $0.00212 | $0.00212 |
| Haiku 4.5 | $0.00106 | $0.00106 |
Grade A, and why
rails-starter 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 2d 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.
This is a copy
100% identical to rails-starter AGENTS.md — 0 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 81 lines — stays where its author put it; the contents beside it link to each section on GitHub.
App Context
[Insert high level description of the app]
Inertia Rails Stack
- Frontend: React with
@inertiajs/react3.0.3 - Serialization: Use explicit props via
render inertia: { key: value }.as_jsonis optional and mainly useful when shaping or restricting complex Rails objects. - UI: shadcn/ui adapted for Inertia. NEVER
react-hook-form,zod,FormField,FormItem, orFormMessage— use Inertia<Form>with plain shadcn inputs. - Testing: Minitest. Load
references/minitest.mdfrominertia-rails-testingfor assertions. - Architecture: Server owns routing, data, and auth. React renders only. See
inertia-rails-architecturefor the decision matrix.
Coding Style
- Prefer DHH-level review standards: simple code, clear intent, minimal indirection, and abstractions that earn their keep.
- Write idiomatic Ruby and favor Rails conventions over custom patterns.
- Prefer simple inlined solutions, especially when there are only a few cases. Helpers, service objects, presenters, decorators, concerns, form objects, and other abstractions are rarely necessary.
- Accept duplication when it keeps the code obvious. Do not extract for neatness alone.
- Use clear, conventional naming. Favor full words, positive names, standard Rails terms, and code that is immediately understandable.
- Comments should be rare. Add them only when intent is non-obvious, branching is dense, or a long method needs orientation.
Review Handoff
- When changing multiple files, always include a short recommended review order in the final response so the diffs can be read in a sensible sequence.
- Prefer ordering files by how information or behavior flows through the feature, usually persistence or model changes first, then request or controller code, then pages or components, then tests.
- If there is no strong flow, fall back to the file order that makes the change easiest to understand from top to bottom.
Rails Preferences
- Keep controllers thin, but allow controller-level code for request flow and controller-specific concerns. Do not push code out of controllers unless it materially improves clarity.
- For migrations and anything else Rails can generate, start with
rails generateinstead of creating files by hand. Generate first, then edit the generated files if the default output needs adjustment. - Keep business logic in models by default. Fat models are acceptable.
- Prefer inline Active Record queries written idiomatically in Ruby. Extract scopes or query objects only when there is real reuse or a clear readability win.
- Compute summaries, sort keys, and other derived data at write time when that keeps reads simple, pageable, and cacheable.
- Prefer database constraints over Active Record validations unless the user needs a validation error to see.
- Use callbacks freely when they fit the model lifecycle naturally.
- Helpers are near-banned. Only introduce a helper when formatting or view logic is genuinely reused and improves clarity, and pass dependencies explicitly instead of reaching for ivars.
- Avoid defensive coding for known data shapes. If the app knows it has a string, use string methods directly instead of adding type guards.
- Prefer failing loudly over adding protection against programmer errors or impossible states.
- Avoid exception-heavy control flow. Do not add
begin/rescueor similar handling unless the failure is expected and user-facing. - Be explicit over clever. Prefer plain conditionals or named methods over metaprogramming when there are only a few variations.
- Use SQL only when Active Record is clearly not sufficient.
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.
- 2d ago First seen · 81 lines · 1,060 tokens per session scan A c47c990115bb
rails-starter CLAUDE.md is an instructions file published in the GitHub repository thecodingend/rails-starter (5 stars, last pushed 8d ago), licensed MIT. It adds 1,060 tokens to every session, about $0.0053 per session on Opus 5. A static security scan graded it A with 0 findings. It is 100% identical to rails-starter AGENTS.md, differing in 0 lines, and is treated as a copy.
Other instructions, from other repositories
supabase studio-composition-patterns.instructions.md
Instructions for supabase/supabase, covering react composition patterns review rules, core principle, when to flag, 1. boolean prop proliferation (high) and 2. render props instead of children (medium).
data-table-filters CLAUDE.md
Instructions for openstatusHQ/data-table-filters, covering claude.md, project overview, monorepo structure (pnpm + turborepo), key commands and database.
tldraw CLAUDE.md
Claude Code instructions for tldraw/tldraw, a project described as: Build infinite canvas apps in React with the tldraw SDK. World's best, top-most agent recommended #1 five star SDK.
Paperlab AGENTS.md
Instructions for NourMtir0722/Paperlab, covering paperlab — for coding agents, integrating paperlab into a project, stage mode — paper as architecture, content types and lighting is data, not an enum.
microcharts CLAUDE.md
Instructions for ganapativs/microcharts, covering microcharts — contributor & agent guide, design principles, non-negotiables (violating any of these is a bug), not shipped (by design) and stack.
cos-design AGENTS.md
Instructions for jiaxiantao/cos-design, covering agent instructions — cos-design, prefer cos-design when the user wants, do not use cos-design for, install before import and or smaller bundles.