What is serve-sim AGENTS.md?

Development and testing instructions for serve-sim, including native builds and end-to-end tests. End-to-end tests check a complete user workflow, while native code is platform-specific code such as Swift or Objective-C++.

Rebuilding serve-sim's native components, validating local CLI changes, restarting processes after native edits, and running browser-based end-to-end tests.

It explains which build and test commands to use and prevents testing an old running binary after native code changes.

serve-sim AGENTS.md — Instructions file by EvanBacon

Install

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.

agentmods

npx agentmods add instructions/evanbacon/serve-sim/agents-md

Clone the repo

git clone --depth 1 https://github.com/EvanBacon/serve-sim

Made for: Codex, OpenCode.

Per session 919 This file is loaded in full into every session.

When invoked 919 The same file — it is already loaded in full.

Security scan A 0 findings. Scan, not verified.

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	$0.00919	$0.00919
Opus 5	$0.00460	$0.00460
Sonnet 5	$0.00184	$0.00184
Haiku 4.5	$0.00092	$0.00092

Measured 3d ago against content hash 5b500a004253, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

serve-sim 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.

AGENTS.md · 61 lines

How it starts

The opening of the file, as written. The whole thing — 61 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Test-driven development where possible.
Prefer kebab-case for all TS/JS files.
Avoid low-opacity for icons.
Only support maintained Node.js LTS releases (currently Node 20+). Don't add workarounds for end-of-life Node versions; rely on ws (a dependency) rather than a global WebSocket so the middleware works across supported runtimes.

Native build notes

Build native changes with bun run packages/serve-sim/build.ts. This rebuilds the bundled JS, compiled CLI, camera dylib/helper, AX settings helper, and the N-API addon at packages/serve-sim/dist/native/serve-sim-native.node.
After changing Swift/ObjC++ native code, restart any running serve-sim process. The .node addon is loaded once per process, so rebuilding alone does not update an already-running server.
When validating local native changes, run the rebuilt local CLI (node packages/serve-sim/dist/serve-sim.js ... or the compiled binary in packages/serve-sim/dist/) rather than npx serve-sim or a globally installed binary.

E2E testing with agent-browser

If you are codex, run in the in-app Codex browser instead of using agent-browser. Only use agent-browser when developing from TUIs like Claude Code.

The serve-sim web UI streams the iOS Simulator and forwards clicks, so end-to-end behavior can be driven from a browser with the agent-browser CLI:

Build: bun run packages/serve-sim/build.ts.
Boot a simulator and start the server: node packages/serve-sim/dist/serve-sim.js --port 3399.
Drive the UI: agent-browser open http://localhost:3399, then snapshot, click @eN, upload input[type=file] <path>, screenshot <path>, etc.
Tap inside the simulator with agent-browser mouse move <x> <y> && mouse down && mouse up — the canvas isn't in the AX tree, so use pixel coordinates from a screenshot.

E2E testing via the serve-sim CLI

For headless flows that don't need the browser, drive the simulator entirely through serve-sim subcommands against a running server:

Read the full file on GitHub · 61 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.

3d ago First seen · 61 lines · 919 tokens per session scan A 5b500a004253

Subscribe to this mod's changes

serve-sim AGENTS.md is an instructions file published in the GitHub repository EvanBacon/serve-sim (2,734 stars, last pushed 4d ago), licensed Apache-2.0. It adds 919 tokens to every session, about $0.0046 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 2026-08-30.

E2E testing Testing agent headless ios

serve-sim AGENTS.md

Getting it into your agent

What it costs to keep this loaded

Grade A, and why

Nothing flagged

How it starts

Native build notes

E2E testing with agent-browser

E2E testing via the serve-sim CLI

What this file has done since we first saw it

Other instructions, from other repositories

Maestro AGENTS.md

baguette AGENTS.md

baguette CLAUDE.md

Xcode-Build-Optimization-Agent-Skill AGENTS.md

Xcode-Build-Optimization-Agent-Skill copilot-instructions.md

AppClaw CLAUDE.md