Getting it into your agent
There is no command for this one: it runs only inside a plugin, and the catalogue could not identify which plugin ships it. The source is linked below.
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.
[](https://agentmods.dev/instructions/kobiton/automate/claude-md)<a href="https://agentmods.dev/instructions/kobiton/automate/claude-md"><img src="https://agentmods.dev/badge/instructions/kobiton/automate/claude-md.svg" alt="Measured on agentmods" 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 | $0.05785 | $0.05785 |
| Opus 5 | $0.02892 | $0.02892 |
| Sonnet 5 | $0.01157 | $0.01157 |
| Haiku 4.5 | $0.00579 | $0.00579 |
Grade B, and why
automate CLAUDE.md scanned grade B with 1 finding 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 6d 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.
Reads MCP configurationmediumAgent snooping
mcp.json carries server URLs and auth tokens; reading it lets a mod discover and abuse other integrations.
| Claude Code | `.mcp.json` (OAuth) + `.mcp.apikey-example.json` (API key) | `.claude-plugin/plugin.json` + `.claude-plugin/marketplace.json` | `skills/*/SKILL.md` | How it starts
The opening of the file, as written. The whole thing — 211 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
This file provides guidance to Claude Code when working with code in this repository.
Getting started
The user journey the plugin serves, end to end: /automate:setup → list/reserve a device → drive it (one skill, below) → save the session as a test case (saveTestCase) → optionally create and monitor a test run.
Every skill is a step on that path; route by intent:
| Reach for this when the user wants… | Skill |
|---|---|
| to run local Appium scripts (Node.js/Python/.NET/Java) on Kobiton devices | run-automation-suite |
| a clean hands-off run of a described flow, saveable as a test case | drive-automation-session |
| quick inspection / troubleshooting — poke at a device, pull logs, push files | run-interactive-session |
| to kick off a test run from a test case or suite | create-test-run |
| to watch a running test run and catch blockers | monitor-test-run |
Vocabulary the skills assume: session (one recorded device connection), test case (saved replayable steps, usually from a session), test run (execution of a case/suite across devices), test suite (ordered collection of cases), reservation (exclusive device hold), device UDID (unique device identifier), live remediation (browser takeover to fix a blocked execution mid-run).
Ambiguous prompts ("test the login screen of app ABC" — goal named, method not): route by signal — script files mentioned → run-automation-suite; flow described / wants it repeatable → drive-automation-session; diagnostic verbs (explore, poke, logs, adb, push/pull) → run-interactive-session; names an existing case/suite → create-test-run; run already in flight → monitor-test-run.
With no deciding signal, ask ONE question (scripts / agent-driven / hands-on) and recommend drive-automation-session — any platform, plain-language goal, and the session stays saveable via saveTestCase.
Never silently pick run-interactive-session for a "test X" goal: CLI sessions don't feed saveTestCase, so that route forfeits the saveable outcome.
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.
- 6d ago First seen · 211 lines · 5,785 tokens per session scan B 784d02f180ec
automate CLAUDE.md is an instructions file published in the GitHub repository kobiton/automate (12 stars, last pushed 2d ago), licensed MIT. It adds 5,785 tokens to every session, about $0.0289 per session on Opus 5. A static security scan graded it B with 1 finding (reads mcp configuration). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
skillsmith AGENTS.md
AGENTS.md instructions for ApexIQ/skillsmith, covering agents.md, 1. prime directives, 2. the 7-stage development cycle, 3. execution standard and 4. slash command registry (mcp mapping).
skillsmith CLAUDE.md
Claude Code instructions for ApexIQ/skillsmith, covering claude.md, prime directives, 7-stage workflow, strategic branching (brain-aware tooling) and slash commands.
skillsmith GEMINI.md
Gemini CLI instructions for ApexIQ/skillsmith, covering gemini.md, prime directives, 7-stage workflow, execution policy and memory and cost policy.
skillsmith copilot-instructions.md
Copilot instructions for ApexIQ/skillsmith, covering github copilot instructions, project overview, required context and workflow.
sumulige-claude AGENTS.md
AGENTS.md instructions for sumulige/sumulige-claude, covering agents.md - agent 系统详解, technical co-founder working agreement (always-on), layer 1: 什么是 agent？ (5 分钟), 最简单的解释 and 类比：餐厅团队.
codexspec CLAUDE.md
Claude Code instructions for Zts0hg/codexspec, covering claude.md - codexspec development guide, project overview, purpose, technology stack and architecture.