What is memex CLAUDE.md?

Development instructions for Memex, a project that makes an AI assistant's conversation context available to Claude Code. Its storage, search, capture, and MCP components provide the supporting infrastructure.

Use it when developing or maintaining Memex, especially when changing its architecture, roadmap, documentation, storage, search, or context-capture features.

It keeps project decisions and context from being lost between sessions or contributors. It also sets rules for planning, reviewing code, updating project documents, and checking for security problems.

How dioniipereyraa/memex configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to dioniipereyraa/memex. 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/dioniipereyraa/memex/main/CLAUDE.md

Clone the repo

git clone --depth 1 https://github.com/dioniipereyraa/memex

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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/dioniipereyraa/memex/claude-md/github.svg)](https://agentmods.dev/instructions/dioniipereyraa/memex/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/dioniipereyraa/memex/claude-md"><img src="https://agentmods.dev/badge/instructions/dioniipereyraa/memex/claude-md/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/instructions/dioniipereyraa/memex/claude-md"><img src="https://agentmods.dev/badge/instructions/dioniipereyraa/memex/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 20,569 This file is loaded in full into every session.

When invoked 20,569 The same file — it is already loaded in full.

Security scan C 3 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.

Model	Per session	Once invoked
Fable 5.1	$0.20569	$0.20569
Opus 5.5	$0.08228	$0.08228
Sonnet 5.5	$0.04114	$0.04114
Haiku 4.5	$0.02057	$0.02057

Measured 2mo ago against content hash 1fb94fda8202, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade C, and why

memex CLAUDE.md scanned grade C with 3 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 2mo 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.

Downloads and executes remote codehighSupply chain

curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.

- **A `curl ... | sh` installer runs under the system `/bin/sh`, not its shebang.** `install-pypi.sh` declared `#!/usr/bin/env bash` and used `set -euo pipefail`, but the documented one-liner pipes it to `sh`. On Debian/

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

- **A request that carries embeddings must be batched by serialized size, never by a fixed item count.** `memex sync push`/`reconcile` batched conversations 500-at-a-time, bounded only by the server's `_MAX_SYNC_UUIDS` c

Runs shell commandslowCapability

Expected in a hook, worth knowing in a rule or an instructions file.

- **A `curl ... | sh` installer runs under the system `/bin/sh`, not its shebang.** `install-pypi.sh` declared `#!/usr/bin/env bash` and used `set -euo pipefail`, but the documented one-liner pipes it to `sh`. On Debian/

CLAUDE.md · 290 lines

How it starts

The opening of the file, as written. The whole thing — 290 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

Context and rules for any Claude Code instance working in this repo (including parallel worktrees).

Project idea in one line

The context Claude.ai has should also be available to Claude Code. Everything else (storage, embeddings, MCP, capture) is plumbing to get there.

Full detail in README.md and ROADMAP.md.

Working rules (apply ALWAYS)

Read the code before and after editing. Before so you do not break anything, after to verify what ended up there.
Keep README, ROADMAP, and DEVLOG in sync with every relevant change. Update them in the same iteration as the code.
Review the code you just wrote for bugs before closing the task.
When closing each ROADMAP phase, audit the whole project for bugs, obsolete code, and vulnerabilities. Deliver a written report.
Plan before coding. No writing code without a clear plan.
If there are real doubts, ask. Do not assume.
Code and plans designed to scale. Clear separation of responsibilities (pure core, swappable transport, embedder and storage behind interfaces).
No em dashes as connectors. Use commas, periods, parentheses. Applies to docs, commits, code, and replies to the user.
No Claude shoutouts in commits. No Co-Authored-By, no AI footers. Commits signed only by the human author.
Apply these rules in every iteration.

Stack

Python 3.12+, package manager uv.
FastMCP for the MCP server (supports stdio and SSE/HTTP).
SQLite + sqlite-vec for storage and vector search.
fastembed by default (zero-config, embedded ONNX) or optional Ollama with nomic-embed-text. Backend configurable via MEMEX_EMBED_BACKEND.
pydantic + pydantic-settings for config and models.
typer + rich for CLI.
pytest, ruff, mypy for test/lint/typecheck.

Read the full file on GitHub · 290 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.

2mo ago First seen · 290 lines · 20,569 tokens per session scan C 1fb94fda8202

Subscribe to this mod's changes

memex CLAUDE.md is an instructions file published in the GitHub repository dioniipereyraa/memex (5 stars, last pushed 3mo ago), licensed MIT. It adds 20,569 tokens to every session, about $0.0823 per session on Opus 5.5. A static security scan graded it C with 3 findings (downloads and executes remote code, makes network calls, runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Memory and context

memex CLAUDE.md next to 6 instructions from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
memex CLAUDE.mddioniipereyraa/memex	Instructions file	5	3mo ago	MIT	C	20,569	Claude Code
vscode buildNext.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	9,719	GitHub Copilot
next.js AGENTS.mdvercel/next.js	Instructions file	143k	yesterday	MIT	A	7,461	Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchain	Instructions file	149k	yesterday	MIT	A	4,680	Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	5,516	GitHub Copilot
vscode copilot-instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	3,093	GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness	Instructions file	274k	2d ago	MIT	A	4,313	Codex, OpenCode

memex: Instructions file for Claude Code