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/pinkpixel-dev/imaginate-mcp/agents-mdgit clone --depth 1 https://github.com/pinkpixel-dev/imaginate-mcpWhat 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 | $0.04591 | $0.04591 |
| Opus 5 | $0.02295 | $0.02295 |
| Sonnet 5 | $0.00918 | $0.00918 |
| Haiku 4.5 | $0.00459 | $0.00459 |
Grade A, and why
imaginate-mcp 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 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.
How it starts
The opening of the file, as written. The whole thing — 355 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Important Rules
- Ask, don't assume. If something is unclear, ask before writing a single line. Never make silent assumptions about intent, architecture, or requirements.
- Simplest solution first. Always implement the simplest thing that could work. Do not add abstractions or flexibility that weren't explicitly requested.
- Don't touch unrelated code. If a file or function is not directly part of the current task, do not modify it, even if you think it could be improved.
- Flag uncertainty explicitly. If you are not confident about an approach or technical detail, say so before proceeding. Confidence without certainty causes more damage than admitting a gap.
- Do NOT change files unless you fully understand the project structure and intent.
- Use a friendly tone and sound engaged with the project.
- Always update the
CHANGELOG.mdandOVERVIEW.mdafter completing a task. Maintain all other documentation as needed or as instructed. Check the system date/time if necessary before updatingCHANGELOG.md. - Use all available agents, skills, MCP, and any other tools/knowledge documents you have access to autonomously as needed. Never assume. If the information or tools required to get the information are available, use them.
- Use web search and/or Context7 tools for up-to-date framework/API documentation. DO NOT ever assume you have all of the information needed if you are unfamiliar with the required packages or the using latest, new versions.
- Always produce modern, elegant, and stylized solutions — avoid outdated or basic implementations.
- No mocks, placeholders, stubs, or temporary implementations remain.
- No files allowed above 500 lines, always run checks after making new files, keep things modular when you edit or create. Refactor when needed, ensure exact same form and function, you are simply making things modular. The 500 line limit does not apply to documentation files.
- Write real, useful tests when the change warrants them. Do not run tests after every tiny fix, wording tweak, style adjustment, or small feature addition. Run only the smallest relevant test set needed to validate substantial code changes, risky logic changes, regressions, or larger features. Do not run the full test suite unless the scope of the change genuinely justifies it.
- Use your "Chrome" plugin or browser verification only after substantial UI work, large scaffolds, major feature additions, broad layout changes, or a very large task list where end-to-end verification is useful. Do not launch Playwright, Chrome, or browser checks after every small fix, edit, style tweak, or minor feature. When browser verification is warranted, ensure you are connecting to the correct host:port and check the relevant desktop, mobile, theme, accessibility, keyboard, tap, touch, and mouse behavior as appropriate. If the server is not running, start it, verify the work, then shut it down.
- Always remember this project is used on mobile heavily. We need to always ensure we design in a mobile friendly way, proper UX, add standard accesability, tooltips, we should always be able to fully control things from tap, touch, mouse, keyboard.
- Always create an Apache 2.0
LICENSEif no license file exists.
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 First seen · 355 lines · 4,591 tokens per session scan A 61eb0df82476
imaginate-mcp AGENTS.md is an instructions file published in the GitHub repository pinkpixel-dev/imaginate-mcp (1 stars, last pushed 7d ago), licensed Apache-2.0. It adds 4,591 tokens to every session, about $0.0230 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
gpt_image_playground AGENTS.md
AGENTS.md instructions for CookSleep/gpt_image_playground, covering agent instructions for gpt-image-playground, 项目概况, 常用命令, 代码风格(强制) and 简单优先.
NanoBananaMCP CLAUDE.md
Claude Code instructions for AceDataCloud/NanoBananaMCP, covering nanobananamcp, project structure, sync from docs and development.
mcp-image-gen CLAUDE.md
Instructions for kevinten-ai/mcp-image-gen, covering claude.md — mcp-image-gen, project overview, architecture, key design decisions and development.
kie-media AGENTS.md
Instructions for step-into-ai/kie-media, covering kie media agent instructions, purpose, bootstrap, agent entry points and safety contract.
limner CLAUDE.md
Claude Code instructions for vinsonconsulting/limner, covering claude.md — limner and storage rules (vinson standard).
kie-media copilot-instructions.md
Instructions for step-into-ai/kie-media: Read and follow the root AGENTS.md for repository development, testing, and paid-task safety.