archcodex CLAUDE.md

Project instructions for ArchCodex, a system that records architectural rules and relationships between parts of a codebase.

In plain words
What is it for?
Use them when modifying an ArchCodex-managed project to load context, read file-specific constraints, follow the required change order, and run the configured checks.
Why use it?
They require the agent to inspect those rules before editing code and validate the result afterward, reducing changes that break project boundaries.

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/archcodexorg/archcodex/claude-md
Clone the repo
git clone --depth 1 https://github.com/ArchCodexOrg/archcodex
Per session 2,323 This file is loaded in full into every session.
When invoked 2,323 The same file — it is already loaded in full.
Security scan A 0 findings. 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 $0.02323 $0.02323
Opus 5 $0.01162 $0.01162
Sonnet 5 $0.00465 $0.00465
Haiku 4.5 $0.00232 $0.00232

Measured yesterday against content hash 465c68e0833e, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

archcodex 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 yesterday.

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 · 278 lines

How it starts

The opening of the file, as written. The whole thing — 278 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Claude Code Instructions for ArchCodex

This project uses ArchCodex for architectural constraints.

BEFORE MODIFYING CODE

STOP. Call this MCP tool first:

archcodex_context { "projectRoot": "/absolute/path/to/project", "module": "src/path/to/module/" }

projectRoot must be an absolute path where .arch/ exists (contains the SQLite database).

This gives you everything in one call:

  • Project Rules - Layer hierarchy + shared constraints across all architectures
  • Modification Order - DEFINES → IMPLEMENTS → ORCHESTRATES
  • 🔴N Impact - How many files break if you change this one
  • Layer Boundaries - CAN/CANNOT import lists
  • Entity Schemas - Fields, relationships, behaviors (inline)
  • Architecture Constraints - forbid, patterns, require, all hints
  • Validation Command - Run after changes

Workflow

1. archcodex_context       ← GET FULL CONTEXT FIRST (module or entity)
2. Read tool               ← READ files (code content)
   archcodex_read          ← GET per-file constraints (auto-injected via pre-read hook)
3. Write code              ← Follow modification order, respect constraints
4. archcodex_check         ← VALIDATE AFTER CHANGES (auto-run via post-write hook)

Note: archcodex_read returns only architectural data (constraints, boundaries, hints) — never file content. Use the agent's native Read tool for code content. The pre-read hook auto-injects constraints when reading src/ files.

MCP Tools

When Tool Call
Before modifying archcodex_context { "projectRoot": "/path", "module": "src/dir/" }
Before creating archcodex_discover { "projectRoot": "/path", "query": "description" }
After changes archcodex_check { "projectRoot": "/path", "files": ["src/**/*.ts"] }
Entity work archcodex_context { "projectRoot": "/path", "entity": "EntityName" }
Task guidance archcodex_action { "projectRoot": "/path", "query": "add order action" }
Verify wiring archcodex_feature_audit { "projectRoot": "/path", "mutation": "duplicateOrder", "entity": "orders" }
Scaffold spec archcodex_spec_scaffold_touchpoints { "projectRoot": "/path", "specId": "spec.orders.duplicateOrder", "entity": "orders" }

Read the full file on GitHub · 278 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. yesterday First seen · 278 lines · 2,323 tokens per session scan A 465c68e0833e

Subscribe to this mod's changes

archcodex CLAUDE.md is an instructions file published in the GitHub repository ArchCodexOrg/archcodex (5 stars, last pushed 3mo ago), licensed MIT. It adds 2,323 tokens to every session, about $0.0116 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.