Borrowing it
Nothing to install: this file belongs to stereosam/stereo-vibe-public. 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/stereosam/stereo-vibe-public/main/AGENTS.mdgit clone --depth 1 https://github.com/stereosam/stereo-vibe-publicWrote 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/stereosam/stereo-vibe-public/agents-md)<a href="https://agentmods.dev/instructions/stereosam/stereo-vibe-public/agents-md"><img src="https://agentmods.dev/badge/instructions/stereosam/stereo-vibe-public/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/stereosam/stereo-vibe-public/agents-md"><img src="https://agentmods.dev/badge/instructions/stereosam/stereo-vibe-public/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 | $0.00758 | $0.00758 |
| Opus 5.5 | $0.00303 | $0.00303 |
| Sonnet 5.5 | $0.00152 | $0.00152 |
| Haiku 4.5 | $0.00076 | $0.00076 |
Grade C, and why
stereo-vibe-public AGENTS.md scanned grade C with 2 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 24d 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.
Downloads and executes remote codehighSupply chain
curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.
- **No download-and-execute.** No `irm | iex`, no `curl | bash` of our own, no encoded Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- **No download-and-execute.** No `irm | iex`, no `curl | bash` of our own, no encoded What it actually says
AGENTS.md
Things that are not visible from the code.
windows/andmacos/are build output. They are byte-for-byte the contents of the published StereoVibe Lite zips. The build script (kept with the private workshop kit) normalizes encodings, copiestemplates/into each package and refuses to build if a check fails. Treat a hand edit here as a patch to be carried back, not as the source.- Two implementations, no shared core.
windows/_не трогать/check-work-tools.ps1andmacos/_не трогать/check-work-tools.share hand-kept at feature parity. A change to one is incomplete until it is made in the other. - Encodings are load-bearing.
.cmdlaunchers are ASCII + CRLF:cmd.exeparses them in the OEM codepage, and one Cyrillic character breaks the launcher with an unrelated-looking error. The Windows.ps1is UTF-8 with BOM + CRLF, or Windows PowerShell 5.1 reads it as ANSI. The macOS script is UTF-8 without BOM + LF, starts with#!/bin/bash..gitattributesstores both package folders as-is — do not "fix" it. - Windows PowerShell 5.1, not 7. Appx cmdlets used for the desktop-app checks fail under PowerShell 7; the launcher forces 5.1 by full path.
- macOS still ships bash 3.2. No associative arrays, no
${var^^}, nomapfile, nocaseinside$( … ). - No download-and-execute. No
irm | iex, nocurl | bashof our own, no encoded commands that fetch code. The one download path is WinGet repair, and it installs only after a valid Microsoft signature check. Antivirus engines flag the other patterns as malware. - Network contact is fixed and documented. A version check (one field read, nothing
executed) and an anonymous summary built from a fixed whitelist of fields, switched off
with
-NoReport/--no-report. Do not add endpoints, and never put free text, paths, user or machine names into the summary. - Every external call has a time limit. winget/npm/brew run in background jobs with a
timeout and a visible bar; a program's own
--versionis limited to 30 s and a timeout means "did not answer", never "missing". - A check may never abort the run. Wrap it, record a row, continue. The report having every line is a feature, not a formality.
- Memory files are never overwritten, and carry no identity. An existing file is left
alone; a fresh
infra_statusgoes next to it as.new.md. No machine name, user name or IP in it; paths start with~. - The full run really installs software and really elevates (Windows). Do not test it on a machine you care about — the check-only mode exercises the same detection paths.
- There are no tests. Verification is a manual check-only run on Windows and on macOS. The risky areas are the sign-in detectors (they depend on file layouts owned by three other vendors) and WinGet repair.
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.
- 24d ago Changed · +16 lines · +343 tokens per session scan A → C 233aef01436d
- 1mo ago First seen · 26 lines · 415 tokens per session scan A c98a4882f219
stereo-vibe-public AGENTS.md is an instructions file published in the GitHub repository stereosam/stereo-vibe-public (1 stars, last pushed 25d ago), licensed MIT. It adds 758 tokens to every session, about $0.0030 per session on Opus 5.5. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
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).
vscode model-prompts.instructions.md
Model-specific prompt authoring and registry guidelines.