Borrowing it
Nothing to install: this file belongs to kunchenguid/chrome-devtools-axi. 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/kunchenguid/chrome-devtools-axi/main/AGENTS.mdgit clone --depth 1 https://github.com/kunchenguid/chrome-devtools-axiWrote 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/kunchenguid/chrome-devtools-axi/agents-md)<a href="https://agentmods.dev/instructions/kunchenguid/chrome-devtools-axi/agents-md"><img src="https://agentmods.dev/badge/instructions/kunchenguid/chrome-devtools-axi/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/kunchenguid/chrome-devtools-axi/agents-md"><img src="https://agentmods.dev/badge/instructions/kunchenguid/chrome-devtools-axi/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 | $0.01444 | $0.01444 |
| Opus 5.5 | $0.00578 | $0.00578 |
| Sonnet 5.5 | $0.00289 | $0.00289 |
| Haiku 4.5 | $0.00144 | $0.00144 |
Grade A, and why
chrome-devtools-axi 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 29d 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 — 50 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
This file provides guidance to coding agents when working with code in this repository.
chrome-devtools-axi is an agent-ergonomic CLI wrapper around chrome-devtools-mcp. Every invocation is short-lived; state that must survive lives in a detached bridge or under ~/.chrome-devtools-axi/. README How It Works owns the process model. README Configuration owns transport selection, named sessions, and keychain isolation. README Page Management owns page selection.
Invariants
- Node 20+, TypeScript, ESM-only (
"type": "module", module resolutionNode16). Relative imports use.jsextensions even from.tsfiles. ensureBridge(src/client.ts) reuses a bridge only after/health?deep=1. A dead browser is terminated and respawned, never reused as a stale endpoint.- Launch modes (
--isolated/--userDataDir) passKEYCHAIN_ISOLATION_CHROME_ARGS; attach modes omit them.test/keychain-isolation.test.tsowns that split. resolveSessionName(src/sessions.ts) rejects path-traversal, unsafe, and all-dot names. A session isolates only the bridge. Two sessions forced onto one port fail loudly (BRIDGE_PORT_IN_USE_EXIT_CODE, 48); do not exportCHROME_DEVTOOLS_AXI_PORTglobally across sessions.- Page identity is never self-healed.
callToolinjects the last AXIselect_pageid (src/selected-page.ts);list_pagesnever sets it. Missing selection fails with "No page is currently selected". After a reconnect,opencreates a new tab. didMcpPageIdentityChange(src/bridge.ts) andisMissingPageError(src/client.ts) own the positional reconnect and missing-page matchers. Read them before changing either. The page-identity notice holder is percallTool, never module state.clearSelectedPageIdreports a drop only when the read-back shows the id is gone.- File-writing commands go through
callTool/handleCallRequest: MCP roots for the cwd and output path, and an MCPisErroris a failure, not printed success.resolveOutputPath(src/paths.ts) is the chokepoint for any new caller-supplied output file or directory. - UID actions fail loud with
STALE_REFunlessparseUidFresh(src/uid-freshness.ts) sees the persisted generation and zero mutations. Generation and selected-page writes are best-effort (src/generation.ts,src/selected-page.ts): a failed write misses one detection or fails the next call loud, and never hangs. getSessionSnapshotIfRunningnever starts the bridge and degrades an invalid session name to null.ensureBridge/stopBridgestill fail loud on an invalid name.src/version.tsis a leaf (node builtins only). The CLI graph stays free of@modelcontextprotocol/sdk; only the bridge constructs an MCP client, soresolveBridgeScriptandBRIDGE_PORT_IN_USE_EXIT_CODElive insrc/bridge-script.ts.test/version-path.test.tsenforces this and must stay free of wall-clock assertions.resolveBridgeScriptprefers a sibling.tsand falls back to the built.js.- chrome-devtools-mcp
evaluate_scriptinvokes its function payload, so every payload must be callable (callFunctioninsrc/run.ts).test/main.test.tsandtest/run.test.tsexecute the sent source.runprints only the script'sconsole.log(RAW_STDOUT_MARKER/wrapStdoutinsrc/cli.ts). src/skill.tsrenders a discovery stub. CLI help is the source of truth; do not copy CLI instructions into the skill.shouldInstallHooksForExecPath(src/hooks.ts) blocks dev entrypoints such aspnpm run devfrom installing hooks.- Transport selection is
resolveTransport,resolveTransportSpec, andbuildTransportArgsinsrc/bridge.ts.test/bridge.test.tscovers them. pnpm-workspace.yamlenforces a minimum release age;axi-sdk-jsandchrome-devtools-axiare exempt..airlock/lint.shmust use pnpm (nevernpm installornpx).test/airlock-lint.test.tsenforces this.- Some
test/client.test.tscases take a couple of seconds on real SIGTERM/SIGKILL escalation. That is expected.
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.
- 29d ago Changed · -86 lines · -2,810 tokens per session 687fe7057926
- 1mo ago Changed · +1 lines · +84 tokens per session e8710df8ba99
- 1mo ago Changed · +14 lines · +29 tokens per session 3748464d096a
- 1mo ago Changed · -1 lines · +11 tokens per session 0c1ed24f2491
- 1mo ago First seen · 122 lines · 4,130 tokens per session scan A 057e26887deb
chrome-devtools-axi AGENTS.md is an instructions file published in the GitHub repository kunchenguid/chrome-devtools-axi (435 stars, last pushed 8d ago), licensed MIT. It adds 1,444 tokens to every session, about $0.0058 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).
deepseek-harness AGENTS.md
AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.