Borrowing it
Nothing to install: this file belongs to YoanWai/agent-manager. 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/YoanWai/agent-manager/main/AGENTS.mdgit clone --depth 1 https://github.com/YoanWai/agent-managerWrote 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/yoanwai/agent-manager/agents-md)<a href="https://agentmods.dev/instructions/yoanwai/agent-manager/agents-md"><img src="https://agentmods.dev/badge/instructions/yoanwai/agent-manager/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/yoanwai/agent-manager/agents-md"><img src="https://agentmods.dev/badge/instructions/yoanwai/agent-manager/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 | $0.03808 | $0.03808 |
| Opus 5.5 | $0.01523 | $0.01523 |
| Sonnet 5.5 | $0.00762 | $0.00762 |
| Haiku 4.5 | $0.00381 | $0.00381 |
Grade A, and why
agent-manager 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 4d 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 — 277 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent notes
Project context for coding agents. Human setup and PR conventions live in .github/CONTRIBUTING.md, and what a review looks for in REVIEW.md. Fill the Scope section of the pull request template: a review reads it to tell the change you meant from the change you made.
Product boundary
Keep agent-manager a thin wrapper around the supported TUIs. Preserve upstream defaults and user configuration; avoid hardcoded models, version-specific behavior, and recreating upstream settings. Apply the Thin Wrapper Principle to every integration, including shared controls, capability discovery, and status detection. Provider-specific features multiply ongoing maintenance. Supporting a TUI does not mean mirroring its feature set; prefer workspace features shared across tools.
What every change covers
A feature, a tool, or a fix is done when it holds on the whole matrix. The
matrix is every CLI in builtinTools (internal/config/config.go), every
platform .goreleaser.yaml builds plus WSL2, every terminal that reaches
the manager, local or over SSH, and both input methods. A change that
covers a subset names what it leaves out in the PR description and what
covering the rest would take. Disclosure alone does not approve partial tool
coverage. REVIEW.md defines the allowed exceptions.
- Keyboard and mouse. Every action a key performs is reachable with the
mouse, and every clickable surface has a key. A new row, panel, button,
or handle ships with both, and the footer and the
?key map name the key. - Dynamic over hardcoded. Anything that varies by tool, version, or user is discovered at runtime or read from the tool itself: models, thinking modes, capabilities, prompts, screens. A maintained list of values that an upstream release can change is a defect. Where discovery has no stable, documented interface, the feature stays in the underlying TUI.
- The current CLI release. Each CLI is supported at its latest stable
release. None publishes a long-term-support line. When an upstream release
changes a flag, a subcommand, a file layout, or a screen, the adapter moves
to the new behavior and drops the old one, with no branch for older
releases. The same change sets the tool's
minimum_versionininternal/config/builtin.goto the first release it needs, so a spawn on an older CLI opens the update dialog instead of failing. - Configurable in the UI, or not configurable. A setting a user can
change lives in Settings (
s), is stored by the manager, and takes effect from the picker, the way the keybindings and editor rows do. The value lives instate.db, where the manager, the CLI and the MCP server all read it, andagent-manager settingsreads and writes it from a shell. Every key is named inui.SettingSpecswith the values it takes, so the shell refuses what the modal would never write. A value that is a command, a token or a URL stays off the modal and is set from the shell alone, so the modal does not grow a row per knob. A new per-user file, environment variable, or hand-edited block is the wrong shape; put the choice on a Settings row, behindagent-manager settings, or pick a default in code. - Every tool. New shared behavior covers every applicable tool in the
same PR. A field, status rule, or command added to one
[tools.*]block is added to every tool it applies to. A fix to one tool's adapter or support for a new CLI can stay narrow. Any other partial feature needs evidence and an explicit maintainer decision before merge. Reuse shared paths first. A working first adapter or an unchanged fallback for other tools is not coverage of the new behavior. If the contributor cannot or does not plan to cover the other applicable tools, start with an issue. An optional draft PR can provide a reference implementation linked to that issue. Mark it as reference only with its missing coverage. Keep it in draft until coverage is complete or the maintainer accepts an exception. - Every platform. A
runtime.GOOSbranch, a platform-only command, or a path built from one platform's layout needs its counterpart, or the description names the platform it skips. WSL2 is a platform of its own: the binary is the Linux one, but the clipboard, notifications, and host stats go through Windows interop (internal/wsldetects it), so a Linux branch is checked there too. - Every terminal. The manager draws through tmux into whatever terminal
the user runs, and over SSH that terminal is on another machine. A
feature that talks to the terminal (clipboard, links, notifications,
themes, mouse) works on the plain path first, and a branch keyed on
TERM_PROGRAM,TERM, orSSH_CONNECTIONis an addition to that path, never the only way the feature works. A tmux version gate names the version and keeps the feature usable on tmux 3.1, the oldest docs/install.md supports.
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.
- 4d ago Changed · +16 lines · +180 tokens per session ef417a66c70f
- 5d ago Changed · +7 lines · +119 tokens per session ca9c2781b7a4
- 10d ago Changed · +23 lines · +386 tokens per session 0c30ed419420
- 12d ago Changed · +3 lines · +29 tokens per session 020daaa34ce6
- 17d ago Changed · +14 lines · +245 tokens per session d431ebf7d879
- 21d ago Changed · +1 lines · +24 tokens per session 115a915760b9
- 24d ago Changed · +4 lines · +50 tokens per session 8bad01f8ff75
- 1mo ago Changed · +96 lines · +1,485 tokens per session 59140aa652c4
- 1mo ago Changed 70b679b4eb48
- 1mo ago First seen · 113 lines · 1,290 tokens per session scan A 31dea5436b1a
agent-manager AGENTS.md is an instructions file published in the GitHub repository YoanWai/agent-manager (625 stars, last pushed today), licensed Apache-2.0. It adds 3,808 tokens to every session, about $0.0152 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.