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/raytone-lab/agentcanvas/agents-mdgit clone --depth 1 https://github.com/raytone-lab/agentcanvasWhat 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 | $0.01550 | $0.01550 |
| Opus 5 | $0.00775 | $0.00775 |
| Sonnet 5 | $0.00310 | $0.00310 |
| Haiku 4.5 | $0.00155 | $0.00155 |
Grade A, and why
agentcanvas 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 — 203 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Design Source of Truth
Before generating or changing any UI, read DESIGN.md (project root) — the visual + interaction spec (colors, typography, geometry, elevation, component contracts, motion). Use semantic tokens only; no raw hex or raw px motion durations. Component architecture lives in docs/COMPONENT_SYSTEM.md.
Project Mode
This project is AgentUX Scaffold Configurator: a default Agent frontend template plus focused UX presets on top of AgentUX SDK.
Optimize for the shortest path to a working, maintainable scaffold. Challenge goals or paths that drift toward a generic low-code builder, infinite canvas, heavy component assembly surface, or decorative UI prototype without exportable engineering structure.
Core Product Rule
Do not build a Figma-style infinite canvas.
The product should feel configurable, but the implementation must be schema-driven and region-constrained so exported code remains clean. Users should start from a strong default Agent frontend, then adjust key UX presets.
The primary user journey is: users assemble and tune the Agent UI/UX in this app, click download/export, receive a runnable code package, and continue second-stage development of their own Agent frontend from that package. AgentCanvas is the scaffold configurator and package generator, not the long-term hosted runtime for the user's Agent product.
Do not treat export as a cosmetic manifest or preview-only summary. The export path must produce real project files that can be installed, run, edited, and extended outside AgentCanvas.
Allowed regions:
maincomposerright-panelbottom-dockoverlay
Free positioning is not a primary MVP interaction. Optional snap placement is acceptable only for composer, output, Git, and overlay/floating components with clear bounds.
Even without drag/drop as the primary interaction, keep each major UI area independent and copyable. The builder should be able to customize Chat, Composer, Output, Git, Activity UX, Provider/Harness controls, and Debug Dock separately.
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 · 203 lines · 1,550 tokens per session scan A 0b5792ba9174
agentcanvas AGENTS.md is an instructions file published in the GitHub repository raytone-lab/agentcanvas (10 stars, last pushed 3d ago), licensed MIT. It adds 1,550 tokens to every session, about $0.0077 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.
Other instructions, from other repositories
ii-agent AGENTS.md
Instructions for Intelligent-Internet/ii-agent, covering ii-agent contributor guide, quick start, repository map, mandatory rules and architecture.
cyrus copilot-instructions.md
Copilot instructions for cyrusagents/cyrus: Note, there is a need to maintain the use of '--print' when running the claude exec commands because that is what makes it non-interactive.
framework AGENTS.md
Instructions for ai-driven-dev/framework, covering agents.md, behavior, communication, action and memory management.
agnix bad-frontmatter.instructions.md
Instructions for agent-sh/agnix, a project described as: The missing linter and lsp for AI coding assistants. Validate CLAUDE.md, AGENTS.md, SKILL.md, hooks, MCP. Plugin for all major IDEs included, with autofixes.
skills CLAUDE.md
Instructions for capawesome-team/skills, covering claude.md, project overview, repository structure, skill.md format and writing agent-optimized skills.
AgentGo AGENTS.md
Instructions for yeasy/AgentGo, covering agents.md, purpose, startup instructions, trust & safety and failure modes.