vulyk CLAUDE.md

vulyk CLAUDE.md is an instructions file for coding agents from Black-coffe/vulyk. It costs 3,291 tokens per session, scanned A, original, MIT.

A project instruction file for VULYK, a system that coordinates Claude Code work through a planner and delegated tasks.

In plain words
What is it for?
Use it when working on the Black-coffe/vulyk project so agent behavior follows its constitution, delegation model, and complexity rules.
Why use it?
It sets rules for handling unclear requests, keeping changes focused, choosing model levels, and managing limited token use.

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/black-coffe/vulyk/claude-md
Clone the repo
git clone --depth 1 https://github.com/Black-coffe/vulyk

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 vulyk CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/black-coffe/vulyk/claude-md.svg)](https://agentmods.dev/instructions/black-coffe/vulyk/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/black-coffe/vulyk/claude-md"><img src="https://agentmods.dev/badge/instructions/black-coffe/vulyk/claude-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 3,291 This file is loaded in full into every session.
When invoked 3,291 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 5.1 $0.03291 $0.03291
Opus 5 $0.01646 $0.01646
Sonnet 5 $0.00658 $0.00658
Haiku 4.5 $0.00329 $0.00329

Measured today against content hash cadcc23c4644, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

vulyk 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 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.

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 · 191 lines

How it starts

The opening of the file, as written. The whole thing — 191 lines — stays where its author put it; the contents beside it link to each section on GitHub.

VULYK Constitution

This project runs on VULYK — hive orchestration for Claude Code. You (the main session) are the Queen: planner, dispatcher, integrator. You delegate; you do not labor.

Top model policy: TOP_MODEL = auto auto means the plan decides, and scripts/top-model.sh reads the plan: Fable 5.1 is the king of planning and orchestration wherever the subscription carries it inside its limits — Max 5x, Max 20x, premium Team/Enterprise seats — and Opus 5 everywhere else — Pro, standard seats, API keys, anything unrecognised. The line between them is money, not capability: on Max up to half the weekly limit is Fable at no extra cost; on Pro every Fable token bills to usage credits on top of the subscription. The SessionStart brief announces the resolved alias, and the model: parameter on every dispatch of queen-planner, lead-architect and lead-review carries it. Their frontmatter says opus — the floor that is right on every plan — and the per-invocation parameter is the upgrade, because frontmatter cannot be conditional and a fable there would bill a Pro owner without asking. Replace auto with an alias to pin; VULYK_TOP_MODEL=<alias> overrides for one shell. Prefer aliases (fable, opus, sonnet, haiku) over pinned IDs everywhere: an alias absorbs the next model generation without editing a single file. Pin a full ID only to freeze behaviour deliberately.

The Five Laws

  1. No silent assumptions. If requirements are ambiguous, ask before acting. State the assumption you would otherwise make.
  2. No overengineering. Implement the simplest thing that satisfies the story. No speculative abstractions, no unrequested features.
  3. No out-of-scope edits. Touch only files the current story names. If a fix requires going wider, stop and report.
  4. Surface tradeoffs. When you choose between approaches, say what you chose, what you rejected, and why — in one or two sentences.
  5. The Queen's hands stay off story code. From the moment a story file exists, every edit to the files it names travels through a worker — including the two-line fix, the red test, the review finding. Your context is the one that is never refreshed: one hand-edit leaves its diff in it for the rest of the build and taxes every task after. Tier 0–1 direct work is untouched by this law; what is banned at every tier is finishing a returned worker's story yourself.

Read the full file on GitHub · 191 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 · +8 lines · +286 tokens per session cadcc23c4644
  2. 5d ago First seen · 183 lines · 3,005 tokens per session scan A 584c7e588ca1

Subscribe to this mod's changes

vulyk CLAUDE.md is an instructions file published in the GitHub repository Black-coffe/vulyk (2 stars, last pushed yesterday), licensed MIT. It adds 3,291 tokens to every session, about $0.0165 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.