kernel-best-practices

A set of rules for using Kernel, a cloud browser-automation and web-scraping service. It covers browser cleanup, time limits, login sessions, bot detection, and proxy choices.

In plain words
What is it for?
Use it when automating websites or scraping pages with Kernel, especially when authentication or bot detection is involved.
Why use it?
It gives an agent practical safety rules for browser tasks and helps avoid leaving browsers running or exposing credentials in code.

Cursor rule

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 rules/kernel/skills/kernel-best-practices
Clone the repo
git clone --depth 1 https://github.com/kernel/skills
Per session 10 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 225 The whole file, excluding the scripts and references it only reads on demand.
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.

ModelPer sessionOnce invoked
Fable 5 $0.00010 $0.00225
Opus 5 $0.00005 $0.00112
Sonnet 5 $0.00002 $0.00045
Haiku 4.5 $0.00001 $0.00022

Measured yesterday against content hash 8afbec9bea23, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

kernel-best-practices 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 yesterday.

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.

rules/kernel-best-practices.mdc · 15 lines

What it actually says

  • always delete browsers when done. use try/finally to guarantee cleanup.
  • set timeout_seconds on every browser as a safety net. default is 60s — increase for longer tasks, max is 72h.
  • use stealth: true when accessing any site with bot detection. stealth mode adds a recaptcha solver and residential proxy automatically.
  • use browser profiles for sites requiring authentication. set save_profile_changes: true to persist session changes.
  • prefer the playwright execution API (execute_playwright_code) for simple one-off scripts. use the SDK + CDP connection for complex multi-step automations.
  • use headless: true for faster execution when you don't need to watch the browser via live view.
  • proxy quality for anti-detection, best to worst: mobile > residential > ISP > datacenter.
  • no charges for idle time — only for active browser usage.
  • never hardcode credentials in automation code. use kernel's managed auth or browser profiles instead.
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.

  1. yesterday First seen · 15 lines · 10 tokens per session scan A 8afbec9bea23

Subscribe to this mod's changes

kernel-best-practices is a cursor rule published in the GitHub repository kernel/skills (10 stars, last pushed 5d ago), licensed MIT. It adds 10 tokens to every session and 225 once invoked, about $0.0001 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 2026-08-31.