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/ds1/boilerplate.md/claude-mdgit clone --depth 1 https://github.com/ds1/boilerplate.mdWrote 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/ds1/boilerplate.md/claude-md)<a href="https://agentmods.dev/instructions/ds1/boilerplate.md/claude-md"><img src="https://agentmods.dev/badge/instructions/ds1/boilerplate.md/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 | $0.01820 | $0.01820 |
| Opus 5 | $0.00910 | $0.00910 |
| Sonnet 5 | $0.00364 | $0.00364 |
| Haiku 4.5 | $0.00182 | $0.00182 |
Grade B, and why
boilerplate.md CLAUDE.md scanned grade B 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 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
- **Tool permissions** (e.g., "allow rm commands", "allow docker") -> add to `~/.claude/settings.json` under `permissions.allow` How it starts
The opening of the file, as written. The whole thing — 141 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Global Claude Code Instructions
Preference Persistence
When offering Yes/No options with a "during this session" scope, always include a third option to save the preference globally. For example:
1. Yes
2. Yes, always do this (session only)
3. Yes, always do this (save globally)
If the user picks option 3:
- Behavioral preferences (e.g., "always run tests after changes", "don't ask before committing") -> append to this file (
~/.claude/CLAUDE.md) - Tool permissions (e.g., "allow rm commands", "allow docker") -> add to
~/.claude/settings.jsonunderpermissions.allow
Default to recommending option 3 over option 2. Session-only preferences are lost on context compression and /clear, making them unreliable. Prefer durable persistence.
Writing Style
- Never use em dashes (—). Use periods, commas, or restructure the sentence instead.
- Minimize hyphen usage. Prefer spelling out compound modifiers or restructuring.
Safety Net (bypassPermissions mode is ON)
Since all actions are auto-approved, follow these guardrails:
- Announce destructive actions (rm, git reset, drop table, etc.) before executing. Give the user a moment to cancel.
- Commit atomically. Make small, focused commits so any change can be cleanly reverted with
git revert. - Never force-push to main/master without explicit confirmation.
- Never delete files outside the project directory without explicit confirmation.
- Never modify
~/.claude/settings.jsonor~/.claude/CLAUDE.mdwithout showing the proposed change first and getting a yes.
Project State Management
CLAUDE.md is for stable instructions only. Never store TODOs, changelogs, sprint status, session context, or completed work in CLAUDE.md. These change frequently and burn context window on every conversation.
Use this consistent file structure across all projects:
project/
├── CLAUDE.md # Stable: commands, conventions, architecture, gotchas
├── STATUS.md # Active: current TODOs, in-progress work, blockers
├── CHANGELOG.md # History: versioned release notes, completed milestones
├── _planning/ # Planning (optional, for larger projects)
│ ├── backlog.md # Prioritized feature backlog
│ ├── sprint.md # Current sprint scope and progress
│ ├── plans/ # Saved plans from Plan Mode (auto-saved here)
│ │ └── YYYY-MM-DD-plan-name.md
│ └── research/ # Research docs, vendor analysis, etc.
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.
- 6d ago First seen · 141 lines · 1,820 tokens per session scan B 68a915683e7d
boilerplate.md CLAUDE.md is an instructions file published in the GitHub repository ds1/boilerplate.md (5 stars, last pushed 6mo ago), licensed MIT. It adds 1,820 tokens to every session, about $0.0091 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
ai-website-cloner-template copilot-instructions.md
Instructions for JCodesMore/ai-website-cloner-template, covering this is not the next.js you know, website reverse-engineer template, what this is, tech stack and commands.
ai-website-cloner-template AGENTS.md
Instructions for JCodesMore/ai-website-cloner-template, covering this is not the next.js you know, website reverse-engineer template, what this is, tech stack and commands.
django-saas-boilerplate CLAUDE.md
Claude Code instructions for eriktaveras/django-saas-boilerplate, covering django saas boilerplate, tech stack, project structure, key conventions and user model.
appointme CLAUDE.md
Instructions for bravodev-hub/appointme, covering claude.md, build and development commands, running the full stack (recommended), backend only and frontend only.
lovable-boilerplate global.instructions.md
Instructions for chihebnabil/lovable-boilerplate, covering react + vite + shadcn/ui template - premium design system, critical rules - read first, never create, always create and essential commands.
lovable-boilerplate components.instructions.md
Instructions for chihebnabil/lovable-boilerplate, covering component-specific guidelines, component architecture rules, ui components (src/components/ui/), common components (src/components/common/) and feature components (src/components/features/).