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/xiaojiou176-open/notebackup/agents-mdgit clone --depth 1 https://github.com/xiaojiou176-open/notebackupWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/xiaojiou176-open/notebackup/agents-md)<a href="https://agentmods.dev/instructions/xiaojiou176-open/notebackup/agents-md"><img src="https://agentmods.dev/badge/instructions/xiaojiou176-open/notebackup/agents-md.svg" alt="Measured on agentmods" height="20"></a>What 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 | $0.01209 | $0.01209 |
| Opus 5 | $0.00605 | $0.00605 |
| Sonnet 5 | $0.00242 | $0.00242 |
| Haiku 4.5 | $0.00121 | $0.00121 |
Grade A, and why
notebackup 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 3d 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 — 111 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Apple Notes Snapshot Agent Guide
This repository is a local-first Apple Notes export wrapper for macOS.
Mission
Keep the repository reviewable, path-safe, and open-source friendly.
Rules
- Keep changes surgical.
- Prefer fixing repo-owned wrapper code over editing vendored upstream code.
- Keep all repo-owned documentation in English.
- Never let real secrets, tokens, session material, personal email addresses, or live diagnostic residue enter the tracked tree, Git history, release notes, screenshots, fixtures, or public docs.
- Prefer placeholder paths such as
${HOME}/...,/path/to/..., or<local-reference-root>/...in docs, fixtures, and examples instead of real machine-specific absolute paths. - Treat any dirty worktree as meaningful until proven otherwise; preserve code before cleanup, rewrite, branch deletion, or remote cutover.
- Before destructive Git or GitHub actions, record a truth anchor and create a rollback asset such as a bundle, backup branch, or equivalent snapshot.
- Never track
.agents/,.agent/,.codex/,.claude/,.serena/,.runtime-cache/,generated/launchd/,logs/, or runtime log files. AGENTS.mdandCLAUDE.mdmay be tracked when they are part of the public repository contract.- Keep
current tracked tree,Git history, andGitHub control-planestatements separate.
Execution hygiene
- Treat browser sessions, cloned browser profiles, Docker containers, and caches as scarce shared machine resources, not disposable scratch space.
- Treat
.serena/as a local MCP/shared-tool cache surface. Keep it ignored and out of repo-owned janitor governance. - Treat the repo-managed isolated Chrome root as persistent runtime state for this repo. It is excluded from TTL/cap janitor cleanup.
- Before using Chrome / Chromium, inventory the current machine state first: task-owned windows, tabs, and temporary profiles must be attributable to this task.
- If the machine already has more than 6 Chrome/Chromium instances in play, do not open another one. Prefer non-browser evidence paths, wait for cleanup, or use a non-Chrome path that you can keep isolated and short-lived.
- Treat
6as the hard shared-machine cap. Within that cap, still prefer one task-owned browser session when the current proof does not require more. - Do not open or reuse Chrome / Chromium / browser instances that belong to a different repo or another active L1 worker on the same machine.
- Prefer one task-owned browser session at a time; close extra tabs, windows, and temporary profiles as soon as the current proof is captured.
- If a task requires a temporary browser profile, record the path, keep it
repo-scoped under
.runtime-cache/browser/<task-slug>/, and delete it when the task ends. - A single task should default to one task-owned browser instance and the smallest possible number of tabs.
- If the task only needs to confirm login state, do not keep reopening browser instances. Try once or twice, classify the state, and move on.
- If the browser tooling supports background or non-focus execution, prefer it so the task does not steal focus from other active work on the shared machine.
- Do not leave repo-owned Docker containers, images, or caches running unless the current task still actively needs them.
- If a task starts Docker/compose/temp-server resources, those resources must be identifiable as repo-owned and must be removed before closeout unless an explicit preserved reason is recorded.
- Every task-owned branch, worktree, and PR must end in one of three states:
merged to
main, salvaged tomainthen deleted, or closed/deleted as confirmed no-value residue. - GitHub repo collaboration writes are allowed only when the task explicitly includes branch / PR / review / merge / release closeout for this repo.
- Every non-GitHub external control plane stays read-only by default. Do not perform write actions against Search Console, registrars, social platforms, video platforms, or other outside accounts unless the user explicitly re-authorizes that exact write action in the current conversation.
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.
- 3d ago First seen · 111 lines · 1,209 tokens per session scan A 70127b795e78
notebackup AGENTS.md is an instructions file published in the GitHub repository xiaojiou176-open/notebackup (1 stars, last pushed 2mo ago), licensed MIT. It adds 1,209 tokens to every session, about $0.0060 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
apple-notes-mcp CLAUDE.md
Instructions for sweetrb/apple-notes-mcp, covering claude.md - apple notes mcp server, related documentation, critical: backslash escaping, why this matters and examples.
searching-apple-notes AGENTS.md
AGENTS.md instructions for yinanli1917-cloud/searching-apple-notes, covering codex project instructions, apple-notes-mcp, tech stack, directory structure and conventions.
searching-apple-notes CLAUDE.md
Claude Code instructions for yinanli1917-cloud/searching-apple-notes, covering apple-notes-mcp, tech stack, directory structure, conventions and codex harness.
apple-mail-mcp CLAUDE.md
Instructions for sweetrb/apple-mail-mcp, covering claude.md - apple mail mcp server, configuring imap / smtp (when a user asks to set it up), critical: backslash escaping, why this matters and examples.
dsh-config-manager AGENTS.md
Instructions for xiajiajun516/dsh-config-manager, covering agents.md — dsh config manager 仓库协作指南, 🗣️ 语言与交流, 📦 项目概览, 🗂️ 仓库结构与职责 and ui 分层(改动前必读,避免放错层).
openclip AGENTS.md
Instructions for ganeshmshetty/openclip, covering agents.md, build & test, architecture hard rules, logging and extensions.