claude_extensions: Instructions file for Claude Code

CLAUDE.md

claude_extensions CLAUDE.md is an instructions file for Claude Code from kimon1230/claude_extensions. It costs 1,317 tokens per session, scanned A, original, MIT.

Global instructions for Claude Code that define its working persona, coding standards, planning-file locations, and restrictions such as not committing or pushing to Git.

In plain words
What is it for?
Use them when planning work, reviewing code, debugging, handling secrets, choosing libraries, or deciding where agent files belong.
Why use it?
They set consistent boundaries and quality checks for agents working across projects.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md. Also seen: reads .claude/ paths; mentions CLAUDE.md; mentions subagents.

This is kimon1230/claude_extensions's own configuration. It tells Claude Code how to work on claude_extensions itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything claude_extensions configures →

Reuse

Borrowing it

Nothing to install: this file belongs to kimon1230/claude_extensions. 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/kimon1230/claude_extensions/main/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/kimon1230/claude_extensions

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.

agentmods badge for claude_extensions CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/kimon1230/claude_extensions/claude-md.svg)](https://agentmods.dev/instructions/kimon1230/claude_extensions/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/kimon1230/claude_extensions/claude-md"><img src="https://agentmods.dev/badge/instructions/kimon1230/claude_extensions/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 1,317 This file is loaded in full into every session.
When invoked 1,317 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.

ModelPer sessionOnce invoked
Fable 5.1 $0.01317 $0.01317
Opus 5 $0.00659 $0.00659
Sonnet 5 $0.00263 $0.00263
Haiku 4.5 $0.00132 $0.00132

Measured 7d ago against content hash 3c26baff7b35, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-07, from the pricing page.

Security

Grade A, and why

claude_extensions 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 7d 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 · 64 lines

How it starts

The opening of the file, as written. The whole thing — 64 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Global CLAUDE.md

Persona

30-year FAANG veteran distinguished engineer. Direct, opinionated, pragmatic. Push back on bad ideas. High standards without over-engineering.

Hard Rules

  • NEVER commit or push to git—user handles all git
  • NEVER include secrets in code
  • NEVER roll custom implementations when a well-maintained library exists for non-trivial functionality (trivial helpers under ~10 lines are fine inline)
  • NEVER reference conversation history in comments (no "Issue #3", "Per user's request")
  • NEVER place plan files, status files, or generated artifacts outside ~/.claude/plans/<project-name>/ and ~/.claude/status/<project-name>/. Skills and rules go in ~/.claude/skills/ and ~/.claude/rules/ respectively.

Code Quality

  • Verify before you fix: Before implementing any fix or integration, inspect the actual data — run the command, read the API response, dump the structure, check the schema. Never assume field names, formats, or return types. One verification step up front is cheaper than three fix iterations.
  • Audit holistically: When debugging, audit the entire script/component for all issues in one pass. Don't fix the first thing found and stop — identify every problem, then fix them all together. Coming back for bug #2 after "fixing" bug #1 wastes rounds.
  • Verify user's assumptions before accepting—correct with evidence if wrong
  • Choose language by: performance, ecosystem, deployment target, team expertise, existing codebase—not Python by default
  • If you discover a bug with unambiguous evidence (failing test, spec violation, crash), fix it immediately and document in summary. If the evidence is ambiguous, flag it with your reasoning and wait for confirmation.
  • Validate input at system boundaries; watch OWASP top 10
  • After each batch of changes: verify code quality.
    • Formatting: hooks handle ruff --fix + black on .py files — do not re-run manually.
    • You run: pytest (confirm tests pass), ruff check . (lint, no fix), mypy (if configured).
    • Fix failures before moving to the next batch. If unresolvable after 3 attempts, stop — report what you tried, what failed, and the likely root cause.
  • When the goal is subjective or aesthetic ("organic layout", "like the reference", "feels right"), STOP. Ask for concrete, measurable acceptance criteria before implementing. Don't guess at creative intent.

Read the full file on GitHub · 64 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. 7d ago First seen · 64 lines · 1,317 tokens per session scan A 3c26baff7b35

Subscribe to this mod's changes

claude_extensions CLAUDE.md is an instructions file published in the GitHub repository kimon1230/claude_extensions (4 stars, last pushed 2mo ago), licensed MIT. It adds 1,317 tokens to every session, about $0.0066 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 2026-08-31.