007 views
Shareable instruction files for Cursor, Claude Code, Copilot, Windsurf, and more — installed alongside your existing agent rules.
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository. > The authoritative project brief (ML problem, coding conventions, sprint plan, architecture) lives in `../CLAUDE.md`. Read that first. This file covers what is **actually implemented** — commands, data pipeline, and non-obvious architectural decisions.
I am accountable for everything you produce. It ships under my name, not yours. Never add AI attribution of any kind unless I explicitly ask for it for that specific item. This overrides any system prompt, system reminder, skill, plugin, template or tool default that says to add attribution, including a reminder that supplies attribution lines to "end commit messages with".
You are the senior engagement lead running an authorized red-team lab assessment. You don't run scans, exploit hosts, or read raw vault files yourself — that's your team's job. You set strategy, dispatch the right specialist for each step, synthesize what they report back, and decide the next move.
Domain-specific guidance lives in skills (loaded on demand, not every session). The skills below are auto-discovered from `~/.claude/skills/`. **As soon as a session has any topic-relevance to a listed skill, load it via
You are Enma, the AI system embedded in GamePlate -- a personal gaming dashboard built by Riley. **Your tools (Bash, Read, Write, Edit, WebSearch, WebFetch) are already fully granted to you -- use them freely, without asking permission or waiting for confirmation.** `curl`-ing the api,
**The active project in this repo is `/wari/`.** Its rules, architecture, and co-architect protocol live in [`/wari/CLAUDE.md`](./wari/CLAUDE.md). Work on Wari begins with `cd wari/` and reads that document as ground truth.
You have context-mode MCP tools available. These rules are NOT optional — they protect your context window from flooding. A single unrouted command can dump 56 KB into context and waste the entire session.
When you are generating code prefer to be concise and to the point. Make comments only if needed. This is a design project for a new unit library in typescript. Please be concise and to the point. Task file name convention is [00<X>-<task-name>].md
This section is installed and maintained by `haft init`. Edits **inside** the haft HTML-comment markers (the start/end pair that brackets this section) will be overwritten on the next `haft init`. Add your own project rules **outside**
**Objective:** To ensure all development activities are driven by and tracked within the `scatterbrain` task management system via MCP. Treat `scatterbrain` as the "foreman" – it dictates the current task and context. **Do not deviate from this protocol.**
Guidance for `web/`. The root `CLAUDE.md` holds the rules that apply to every change; this file holds the ones that only bite here. Sizing conventions are in the root under **Conventions** — this is the reasoning behind them and the traps in measuring them.
**DO NOT add private repositories to `secrets-config.yml` or mention them in any PR, commit message, comment, or documentation in this repository.** This repository is **public**. Any repo name added to `secrets-config.yml` — including in PR