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/ginzatech/rn-agent-observer/agents-mdgit clone --depth 1 https://github.com/GinzaTech/rn-agent-observerWhat 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 | $0.01222 | $0.01222 |
| Opus 5 | $0.00611 | $0.00611 |
| Sonnet 5 | $0.00244 | $0.00244 |
| Haiku 4.5 | $0.00122 | $0.00122 |
Grade A, and why
rn-agent-observer 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 — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.
RN Agent Observer instructions
Observer runtime cục bộ cho React Native/Expo, built-in target Android; Windows là
host phát triển đã verify. Yêu cầu Node >= 22.12, pnpm 9.6, adb khả dụng.
Lệnh chính
pnpm check=lint -> format:check -> build -> test; chạy sau mọi thay đổi.- Một package:
pnpm --filter @rn-agent-observer/core test(tương tự vớibuild). - Một file test:
pnpm --filter @rn-agent-observer/core test -- <file-hoặc-pattern>(vitest nhận positional filter). pnpm rn-observe,pnpm mcp:check,pnpm mcp:startchạy thẳngpackages/*/dist— phảipnpm buildtrước khi dùng CLI/MCP sau khi sửa source.devtools-exportcần Metro chạy cho đúng app +adb reverse tcp:8081 tcp:8081(hoặcRN_OBSERVER_METRO_URL); không mở React Native DevTools song song. Với app không instrumentation, dùngapp-state/device-networklàm fallback.
Bản đồ workspace
packages/schemas— Zod schemas + shared types, không chứa logic runtime.packages/core— toàn bộ device/runtime logic (ADB, parser, session, diagnosis, comparison). CLI và MCP là adapter mỏng; không đưa logic vào adapter.packages/rn-instrumentation— telemetry development-only gắn vào app được quan sát (route, fetch, render, long JS task).apps/demo-expo— fixture dogfood, app IDdev.rnagentobserver.demo.- Hướng phụ thuộc:
schemas <- core <- cli/mcp-server; demo-expo dùngrn-instrumentation.
Ràng buộc code
- TypeScript strict +
noUncheckedIndexedAccess+exactOptionalPropertyTypes(tsconfig.base.json). - ESLint chạy
--max-warnings=0;no-explicit-anyvàconsistent-type-importslà error. - Prettier: single quote, semicolon, trailing comma
all; format check nằm trongpnpm check, chạypnpm formatđể sửa. - Chỉ dùng
pnpm; không tạo lockfile npm/Yarn/Bun.
Fixture demo (đừng "sửa")
- PerformanceLab chặn JS 100ms một cách có chủ ý làm regression fixture; không tối ưu hóa.
- NetworkLab dùng fixture nội bộ (0/500/2000ms, HTTP 503), không phụ thuộc Internet.
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 · 68 lines · 1,222 tokens per session scan A 728c95e8e337
rn-agent-observer AGENTS.md is an instructions file published in the GitHub repository GinzaTech/rn-agent-observer (0 stars, last pushed 2d ago), licensed Apache-2.0. It adds 1,222 tokens to every session, about $0.0061 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
native-devtools-mcp AGENTS.md
AGENTS.md instructions for vectora-foundry/native-devtools-mcp, covering agent context: native-devtools-mcp, 🧠 core reasoning loop, 🗺️ capabilities matrix (strategy guide), 🛠️ tool definitions & schemas and 1. vision & perception (the "eyes").
flog AGENTS.md
Instructions for shaominngqing/flog, covering agents.md, 🛡️ design rules that outlive the codebase, what this is, build & test commands and architecture.
flog CLAUDE.md
Instructions for shaominngqing/flog, covering claude.md, 🛡️ design rules that outlive the codebase, what this is, build & test commands and architecture.
scrcpy-mcp AGENTS.md
Instructions for JuanCF/scrcpy-mcp, covering agents.md, project overview, project documentation, build/lint/test commands and build the project.
android-agent CLAUDE.md
Instructions for ghost-in-the-droid/android-agent, covering ghost in the droid — agent tool rules, mcp tool: opencamera, mcp tool: speaktext and general rule.
electron-mcp-server CLAUDE.md
Instructions for halilural/electron-mcp-server, covering project guidelines, git, development and releases.