Plannotator is a local browser-based review tool for coding agents that lets people annotate plans, documents, HTML artifacts, code diffs, and pull requests, then send feedback back to the agent. It is used by developers and teams reviewing work from agents across supported coding tools and version-control systems.
Borrowing it
Nothing to install: this file belongs to backnotprop/plannotator. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/backnotprop/plannotator/main/AGENTS.mdgit clone --depth 1 https://github.com/backnotprop/plannotatorWrote 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/backnotprop/plannotator/agents-md)<a href="https://agentmods.dev/instructions/backnotprop/plannotator/agents-md"><img src="https://agentmods.dev/badge/instructions/backnotprop/plannotator/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/instructions/backnotprop/plannotator/agents-md"><img src="https://agentmods.dev/badge/instructions/backnotprop/plannotator/agents-md.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.01194 | $0.01194 |
| Opus 5.5 | $0.00478 | $0.00478 |
| Sonnet 5.5 | $0.00239 | $0.00239 |
| Haiku 4.5 | $0.00119 | $0.00119 |
Grade A, and why
plannotator 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 today.
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 — 38 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Plannotator agent guidance
Plannotator is a plan review UI for Claude Code that intercepts ExitPlanMode via hooks, letting users approve or request changes with annotated feedback. It also provides code review for git diffs and annotation of arbitrary Markdown files.
This file is a map. The full guidance lives in focused guides under docs/agent-guides/. Read the hard rules below, then the guide for the code or behavior you are changing. CLAUDE.md is a symlink to this file; packages/ui/AGENTS.md adds guidance scoped to packages/ui/.
Hard rules (read before touching anything)
- Reusing the document UI in the commercial Workspaces app? Read
packages/ui/README.mdFIRST: an earlier from-scratch rebuild of this UI broke the app and was reverted. Do not rebuild the UI or recreatepackages/document-ui; add a seam to@plannotator/ui(configurePlannotatorUI()), keep Plannotator's app unchanged, and never delete working code until a human confirms parity in the browser. - Two server runtimes change together. An endpoint change goes into both the Bun server (
packages/server/) and the Pi server (apps/pi-extension/server/); shared logic lives inpackages/shared/. Only the Inbox is Bun-only. See Server Runtimes. - Build order: review, then hook, then opencode.
bun run --cwd apps/review build && bun run build:hook && bun run build:opencode; building out of order copies stale HTML. See Build. - Tailwind
@source. A new directory with.tsxfiles that use Tailwind classes needs a matching@sourceentry in the app'sindex.css. - Testing Rules (full list in Testing Rules):
- Tests and dev scripts never touch the real
~/.plannotator, the real Inbox or user config: sandbox under a tempPLANNOTATOR_DATA_DIR, and mutate globals only inside tests with a restore. - Tests never open a browser: set
NO_BROWSERfromtests/helpers/no-browser.ts, neverPLANNOTATOR_BROWSER=none(or another sentinel). - Tests never run the real GitHub CLI, and never signal an Inbox except through
tests/helpers/inbox-process.ts. - Every test guards a behavior that can regress: no copy snapshots of prose, no round-trip prop tests, no implementation echo.
- Tests and dev scripts never touch the real
- No new Inbox keyboard shortcuts without the owner's approval (see the Inbox shortcuts note in Keyboard Shortcuts). Shortcuts elsewhere go through the shortcut registry.
- Git: never rebase or force-push a shared branch; merge main instead.
- Releases: follow
.agents/skills/release/SKILL.md.
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.
- today Changed · -1,133 lines · -43,709 tokens per session 0c69558220bf
- 1mo ago Changed · +20 lines · +4,287 tokens per session df9ec6bac870
- 1mo ago Changed · +59 lines · +3,063 tokens per session 4bc18dd3f441
- 1mo ago Changed · +64 lines · +1,030 tokens per session 3316c063cc53
- 1mo ago Changed · +2 lines · +584 tokens per session c8d2e1ce5db7
- 1mo ago Changed · +62 tokens per session 0802af8582c7
- 1mo ago Changed · +100 lines · +2,733 tokens per session 22efd8a6efaa
- 1mo ago Changed · +110 lines · +2,493 tokens per session 08e84bd23999
- 1mo ago First seen · 816 lines · 30,651 tokens per session scan A 57acd44f1cdb
plannotator AGENTS.md is an instructions file published in the GitHub repository backnotprop/plannotator (9,683 stars, last pushed today), licensed Apache-2.0. It adds 1,194 tokens to every session, about $0.0048 per session on Opus 5.5. 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
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode copilot-instructions.md
Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).
deno copilot-instructions.md
Copilot instructions for denoland/deno, covering deno development guide for github copilot, network access, high level overview, key directories and key files to understand first.