What is Loupe AGENTS.md?

A set of project instructions for Loupe, a macOS tool that inspects and annotates interface elements for accessibility work. It tells a coding agent how the app is organised, built, and checked, including its Swift package boundary and screen-coordinate rules.

Use it when modifying Loupe's macOS app, especially its accessibility inspection, overlays, Swift code, workspace, or package structure. It guides builds and handoffs without being the application itself.

It prevents changes that build the wrong Xcode project, cross package boundaries incorrectly, or place overlays at the wrong screen position. It also records requirements for accessibility permissions, signing, and verification.

How smughead/Loupe configures Codex and OpenCode

Reuse

Borrowing it

Nothing to install: this file belongs to smughead/Loupe. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/smughead/Loupe/main/AGENTS.md

Clone the repo

git clone --depth 1 https://github.com/smughead/Loupe

Made for: Codex, OpenCode.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/smughead/loupe/agents-md/github.svg)](https://agentmods.dev/instructions/smughead/loupe/agents-md)

Your own site

<a href="https://agentmods.dev/instructions/smughead/loupe/agents-md"><img src="https://agentmods.dev/badge/instructions/smughead/loupe/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.

Your own site · 80×15

<a href="https://agentmods.dev/instructions/smughead/loupe/agents-md"><img src="https://agentmods.dev/badge/instructions/smughead/loupe/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 982 This file is loaded in full into every session.

When invoked 982 The same file — it is already loaded in full.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

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	$0.00982	$0.00982
Opus 5.5	$0.00393	$0.00393
Sonnet 5.5	$0.00196	$0.00196
Haiku 4.5	$0.00098	$0.00098

Measured 1mo ago against content hash 4ce5f549503e, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

Loupe 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 1mo 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.

AGENTS.md · 60 lines

How it starts

The opening of the file, as written. The whole thing — 60 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Loupe Project Instructions

These instructions apply to any coding agent (Claude, Codex, or others). CLAUDE.md only imports this file; edit this file, not that one.

Product purpose

Loupe is a native macOS accessibility utility for inspecting and annotating interface elements. Changes must remain dependable in the real app, where macOS Accessibility permission, overlay behavior, signing, and coordinate conversion all affect what the user experiences.

Project boundaries

Open and build Loupe.xcworkspace; do not substitute the .xcodeproj.
Keep the app shell in Loupe/ and feature implementation in LoupePackage/Sources/LoupeFeature/ consistent with the existing workspace and Swift package boundary.
Types consumed by the app target must remain accessible across that package boundary (mark them public, with a public init()).
Accessibility coordinates use a top-left origin while AppKit screen coordinates use a bottom-left origin. Preserve and verify the existing conversion whenever selection, positioning, or overlays are involved.
Package dependencies belong in LoupePackage/Package.swift; build settings belong in the XCConfig files in Config/ (Shared, Debug, Release, Tests) unless the established project structure requires otherwise.

Working agreements

Inspect Git status before working. Preserve unrelated changes and untracked build output; do not clean, delete, pull, or rewrite repository state unless asked.
Treat signing, entitlements, App Sandbox, Accessibility permission, window level, and overlay event handling as sensitive boundaries. Inspect their current behavior before changing them, and do not broaden permissions merely to make a test pass.
Keep the experience native and accessible. Avoid changes that interfere with other apps, trap input, or leave overlays active unexpectedly.
Do not confuse a successful build with successful accessibility behavior. Permission prompts, first-run state, multi-display coordinates, and the actual overlay must be checked when relevant.
Do not turn conversational promises into standing instructions or memory automatically. Propose additions to this file and make them only when the user approves.

Read the full file on GitHub · 60 lines

Changes

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.

1mo ago First seen · 60 lines · 982 tokens per session scan A 4ce5f549503e

Subscribe to this mod's changes

Loupe AGENTS.md is an instructions file published in the GitHub repository smughead/Loupe (112 stars, last pushed 1mo ago), licensed MIT. It adds 982 tokens to every session, about $0.0039 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 2026-09-30.

Planning Writing and docs accessibility ai-agents developer-tools inspector macos swift

Loupe AGENTS.md next to 6 instructions from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
Loupe AGENTS.mdsmughead/Loupe	Instructions file	112	1mo ago	MIT	A	982	Codex, OpenCode
vscode buildNext.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	9,719	GitHub Copilot
next.js AGENTS.mdvercel/next.js	Instructions file	143k	today	MIT	A	7,461	Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	5,516	GitHub Copilot
langchain AGENTS.mdlangchain-ai/langchain	Instructions file	149k	today	MIT	A	4,680	Codex, OpenCode
vscode copilot-instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	3,093	GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness	Instructions file	275k	3d ago	MIT	A	4,313	Codex, OpenCode

Loupe: Instructions file for Codex