memex CLAUDE.md

memex CLAUDE.md is an instructions file for coding agents from timurgaleev/memex. It costs 3,448 tokens per session, scanned B, original, MIT.

A set of operating rules for AI coding agents working in the memex repository. It defines permitted file and infrastructure actions, including when Terraform, AWS commands, commits, and pushes need approval.

In plain words
What is it for?
Use it to guide safe work in the repository, especially Terraform plans, AWS read-only checks, file edits, and Git operations such as committing or pushing.
Why use it?
It reduces the chance that an agent will make destructive infrastructure or repository changes without the user’s knowledge. It also sets boundaries for what the agent may inspect or edit.

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/timurgaleev/memex/claude-md
Clone the repo
git clone --depth 1 https://github.com/timurgaleev/memex

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 memex CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/timurgaleev/memex/claude-md.svg)](https://agentmods.dev/instructions/timurgaleev/memex/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/timurgaleev/memex/claude-md"><img src="https://agentmods.dev/badge/instructions/timurgaleev/memex/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 3,448 This file is loaded in full into every session.
When invoked 3,448 The same file — it is already loaded in full.
Security scan B 1 finding. 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 $0.03448 $0.03448
Opus 5 $0.01724 $0.01724
Sonnet 5 $0.00690 $0.00690
Haiku 4.5 $0.00345 $0.00345

Measured today against content hash 3f59bc588c75, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

memex 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 today.

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.

Asks for rootmediumPrivilege escalation

A mod that escalates privileges can change anything on the machine, not only the project.

- For new timer units: `sudo systemctl start <unit>` succeeds, then `systemctl is-active` reports OK
CLAUDE.md · 261 lines

How it starts

The opening of the file, as written. The whole thing — 261 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md — Rules for AI agents working in this repo

This file is auto-loaded by Claude Code (and other AI coding agents that honor the convention). It defines the safety and scope rules every agent must follow when editing this codebase.

What you are allowed to do

  • Read files and understand the codebase.
  • Edit config, scripts, and docs when explicitly asked.
  • Run terraform plan and make audit to show what would change.
  • Run AWS CLI read-only commands (describe, list, get).
  • Commit and push ONLY when explicitly asked.

What you are NOT allowed to do without explicit confirmation

Destructive infrastructure changes

  • NEVER run terraform apply without showing the plan first AND getting explicit "yes, apply".
  • NEVER destroy or replace running EC2 instances without explicit confirmation.
  • NEVER change availability zones, instance types, or subnet configurations without warning.
  • If a terraform plan shows ANY resource being destroyed or replaced — STOP and ask first.

File and git operations

  • NEVER run git reset --hard or any destructive git command.
  • NEVER commit files that were not explicitly included in the request.
  • NEVER stage and commit deleted files unless explicitly asked to remove them.
  • Disk files and git files are DIFFERENT — do not confuse them.

Scope of work

  • Do ONLY what was asked. If asked to fix a config file, fix that file only.
  • Do NOT remove integrations, features, or code that was not mentioned.
  • Do NOT "clean up" things that look unused — ask first.
  • Do NOT add, remove, or modify terraform resources beyond what was asked.
  • Do NOT add monitoring, alarms, notifications, health checks, dashboards, or any new AWS infrastructure that the user did not explicitly request, even if it would "improve reliability" or "catch the next failure". Discuss the idea in plain text first, agree on the approach, and only then build.

Key principle

Ask before acting on anything irreversible. The cost of one confirmation message is zero. The cost of a destroyed EC2 instance or lost disk data is high.

Read the full file on GitHub · 261 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. today Changed · +7 lines · +149 tokens per session 3f59bc588c75
  2. 4d ago First seen · 254 lines · 3,299 tokens per session scan B 374219a83933

Subscribe to this mod's changes

memex CLAUDE.md is an instructions file published in the GitHub repository timurgaleev/memex (8 stars, last pushed 2d ago), licensed MIT. It adds 3,448 tokens to every session, about $0.0172 per session on Opus 5. A static security scan graded it B with 1 finding (asks for root). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.