What is stokowski CLAUDE.md?

A set of instructions for Stokowski, a Python service that coordinates Claude Code agents through Linear issues. Linear is a project-management tool where issues track work, and Stokowski gives agents isolated workspaces and manages their sessions.

Use it when developing or maintaining Stokowski, especially its Linear integration, workspace setup, agent launching, retries, state reconciliation, dashboard, or terminal interface.

It gives contributors one reference for understanding the service's purpose, architecture, and important design choices. This makes it easier to work on the codebase consistently.

How Sugar-Coffee/stokowski configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to Sugar-Coffee/stokowski. 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/Sugar-Coffee/stokowski/main/CLAUDE.md

Clone the repo

git clone --depth 1 https://github.com/Sugar-Coffee/stokowski

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/sugar-coffee/stokowski/claude-md.svg)](https://agentmods.dev/instructions/sugar-coffee/stokowski/claude-md)

Your own site

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

Per session 7,031 This file is loaded in full into every session.

When invoked 7,031 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.

Model	Per session	Once invoked
Fable 5.1	$0.07031	$0.07031
Opus 5	$0.03515	$0.03515
Sonnet 5	$0.01406	$0.01406
Haiku 4.5	$0.00703	$0.00703

Measured yesterday against content hash b4a74aa4bbe8, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

stokowski 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 · 531 lines

How it starts

The opening of the file, as written. The whole thing — 531 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Stokowski

Claude Code adaptation of OpenAI's Symphony. Orchestrates Claude Code agents via Linear issues.

This file is the single source of truth for contributors. It covers architecture, design decisions, key behaviours, and how to work on the codebase.

What it does

Stokowski is a long-running Python daemon that:

Polls Linear for issues in configured active states
Creates an isolated git-cloned workspace per issue
Launches Claude Code (claude -p) in that workspace
Manages multi-turn sessions via --resume <session_id>
Retries failures with exponential backoff
Reconciles running agents against Linear state changes
Exposes a live web dashboard and terminal UI

The agent prompt, runtime config, and workspace setup all live in workflow.yaml in the operator's directory — not in this codebase.

Package structure

stokowski/
  artifacts.py     Agent evidence: collection, git isolation, cleanup
  config.py        workflow.yaml parser + typed config dataclasses
  events.py        stream-json event parsing -> RunAttempt state
  ledger.py        Append-only run log + approval-rate summary
  linear.py        Linear GraphQL client (httpx async)
  model_catalogue.py  Known models by provider (data, for the studio dropdown)
  models.py        Domain models: Issue, RunAttempt, RetryEntry
  # workflows/*.yaml alongside the config hold one pipeline each
  orchestrator.py  Main poll loop, dispatch, reconciliation, retry
  prompt.py        Three-layer prompt assembly + the agent reporting contract
  report.py        Structured run reports -> rendered Linear comments
  studio.py        Comment-preserving config editing behind the dashboard
  runner.py        Claude Code CLI integration, subprocess lifecycle
  tracking.py      State machine tracking via structured Linear comments
  workspace.py     Per-issue workspace lifecycle and hooks
  web.py           Optional FastAPI dashboard
  main.py          CLI entry point, keyboard handler
  __main__.py      Enables python -m stokowski

Read the full file on GitHub · 531 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.

yesterday Changed · +233 lines · +3,025 tokens per session b4a74aa4bbe8
6d ago First seen · 298 lines · 4,006 tokens per session scan A 77be9d1af206

Subscribe to this mod's changes

stokowski CLAUDE.md is an instructions file published in the GitHub repository Sugar-Coffee/stokowski (115 stars, last pushed 4d ago), licensed Apache-2.0. It adds 7,031 tokens to every session, about $0.0352 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 2026-08-30.

next.js AGENTS.md

AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,296 tokens

codex AGENTS.md

AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.

openai/codex · 5,182 tokens

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

stokowski: Instructions file for Claude Code