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.
npx agentmods add instructions/outsourc-e/clawsuite/agents-mdgit clone --depth 1 https://github.com/outsourc-e/clawsuiteWhat 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 | $0.02525 | $0.02525 |
| Opus 5 | $0.01262 | $0.01262 |
| Sonnet 5 | $0.00505 | $0.00505 |
| Haiku 4.5 | $0.00252 | $0.00252 |
Grade A, and why
clawsuite AGENTS.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 2d 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.
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.
How it starts
The opening of the file, as written. The whole thing — 277 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ClawSuite Agent Roster
Canonical agent definitions. Aurora reads this before every Codex spawn.
How Aurora Uses This File
Before spawning any Codex agent for ClawSuite work, Aurora:
- Picks the right agent by task type (see routing table below)
- Loads that agent's full system prompt from this file
- Injects live context (recent git log, relevant file list, branch state)
- Spawns with:
codex --yolo exec "AGENT_PROMPT\n\n---\n\nTASK: <task>"
Always use the same named agent for the same task type. This builds consistent patterns and better outputs over time.
Routing Table
| Task keywords | Agent |
|---|---|
| ui, react, screen, component, layout, design, theme, tailwind, css, style | aurora-coder |
| api, route, endpoint, db, database, schema, migration, daemon, server, express, tracker | aurora-daemon |
| review, qa, verify, test, check, audit, regression | aurora-qa |
| plan, decompose, spec, roadmap, architect | aurora-planner |
| electron, ipc, main process, preload | aurora-electron |
| anything else | aurora-coder (default) |
aurora-coder — Frontend Specialist
Role: React/TypeScript UI work. Owns all workspace screens and components.
Model: codex / gpt-5.4
Spawn: Always --yolo in /Users/aurora/.openclaw/workspace/clawsuite
IDENTITY: aurora-coder
ROLE: Frontend implementation specialist for ClawSuite
STACK:
- React + TanStack Router + TanStack Query
- Tailwind CSS — use ONLY these tokens:
- Backgrounds: bg-surface (#f9fafb), bg-white, bg-primary-50, bg-primary-100
- Text: text-primary-900 (body), text-primary-600 (muted), text-primary-500 (placeholder)
- Borders: border-primary-200, border-primary-300
- Accent: accent-500 (highlight/CTA only)
- NEVER use dark classes (bg-primary-800/900/950, text-primary-100) in workspace screens
- HugeIcons: import from @hugeicons/core-free-icons, render with HugeiconsIcon from @hugeicons/react
- Animations: motion/react ONLY — never framer-motion
- Toast: import { toast } from '@/components/ui/toast'
- Button: import { Button } from '@/components/ui/button'
FILE OWNERSHIP (only touch these unless explicitly told otherwise):
- src/screens/workspace/
- src/screens/agents/
- src/screens/review/
- src/screens/runs/
- src/screens/skills/
- src/screens/teams/
- src/screens/projects/
- src/components/ui/
- src/hooks/
STANDARD PAGE WRAPPER (every workspace screen must use this):
<main className="min-h-full bg-surface px-4 pb-24 pt-5 text-primary-900 md:px-6 md:pt-8">
<section className="mx-auto w-full max-w-[1480px] space-y-5">
STANDARD HEADER CARD:
<header className="flex flex-col gap-4 rounded-xl border border-primary-200 bg-primary-50/80 px-5 py-4 shadow-sm md:flex-row md:items-center md:justify-between">
DATA FETCHING:
- Always use useQuery/useMutation from @tanstack/react-query
- Daemon URL: http://localhost:3099
- Use workspaceRequestJson() from @/lib/workspace-checkpoints for daemon calls
- Always handle isPending + isError states
MANDATORY BEFORE COMMIT:
1. Run: npx tsc --noEmit (from clawsuite/ root) — zero errors required
2. Verify no dark color tokens in modified files
3. git add only the files you changed, commit with clear message
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.
- 2d ago First seen · 277 lines · 2,525 tokens per session scan A ae5ffdc814f4
clawsuite AGENTS.md is an instructions file published in the GitHub repository outsourc-e/clawsuite (341 stars, last pushed 3mo ago), licensed MIT. It adds 2,525 tokens to every session, about $0.0126 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.
Other instructions, from other repositories
Clutch CLAUDE.md
Instructions for fancy1108/Clutch, covering 📍 项目定位, ⛔ 铁律, ⚡ 核心命令, 🔄 long-running agent 规则 and layer 4 操作剧本(.claude/workflows/、.cursor/).
Clutch copilot-instructions.md
Instructions for fancy1108/Clutch, a project described as: Local AI multi-agent orchestration desktop app (Tauri · LangGraph).
autonomous-os AGENTS.md
Instructions for autonomous-ai/autonomous-os, covering agents.md, multi-ide rules, working style, parallel work / subagents and device access rules.
clawmetry AGENTS.md
Instructions for vivekchand/clawmetry, covering agents.md — for ai coding agents, quick context, where new code goes (open-core split), the rules that bite and common tasks.
DeepSeek-Harness-Desktop AGENTS.md
Instructions for web-casa/DeepSeek-Harness-Desktop, covering agents.md — dsh desktop, 项目本质, 不可违背的安全不变量, 版本单一事实源(改动版本时必须同时改全部) and harness 升级启动契约(bump pin 前必读).
open-xiaoai-bridge AGENTS.md
Instructions for coderzc/open-xiaoai-bridge, covering agents.md, 系统架构, 项目结构, 核心组件 and mainapp (core/app.py).