grok

grok is a command for coding agents from xiaolai/cc-suite. It costs 35 tokens per session (1,332 once invoked), scanned C, original, ISC.

A command for sending a task from Claude to Grok Build, xAI's coding agent, through the Agent Client Protocol (ACP), a standard way for one program to communicate with another agent.

In plain words
What is it for?
Use it to delegate coding tasks to Grok Build, optionally allowing it to call Claude through the shared MCP server.
Why use it?
It provides bounded, trackable delegation instead of requiring you to run Grok separately and copy results back manually. Jobs can be checked, retrieved, continued, or cancelled through cc-suite.

Command

Installs and runs on its own, but its text points at files inside its plugin — anything it tells you to read at a ${CLAUDE_PLUGIN_ROOT} path is only there once the plugin is installed. Installing the plugin gets both.

Part of the cc-suite plugin — 13 skills, 39 commands, 4 agents, 3 hooks shipped together

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 commands/xiaolai/cc-suite/grok
Clone the repo
git clone --depth 1 https://github.com/xiaolai/cc-suite

Or install cc-suite, the plugin that ships this one along with the rest of its 13 skills, 39 commands, 4 agents, 3 hooks.

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.

agentmods badge for grok

README.md
[![agentmods](https://agentmods.dev/badge/commands/xiaolai/cc-suite/grok.svg)](https://agentmods.dev/commands/xiaolai/cc-suite/grok)
Your own site
<a href="https://agentmods.dev/commands/xiaolai/cc-suite/grok"><img src="https://agentmods.dev/badge/commands/xiaolai/cc-suite/grok.svg" alt="Measured on agentmods" height="20"></a>
Per session 35 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,332 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 2 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 $0.00035 $0.01332
Opus 5 $0.00017 $0.00666
Sonnet 5 $0.00007 $0.00266
Haiku 4.5 $0.00003 $0.00133

Measured 3d ago against content hash 7430d04b62cf, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

grok scanned grade C with 2 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.

Downloads and executes remote codehighSupply chain

curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.

> an install hint (`curl -fsSL https://x.ai/cli/install.sh | bash`).

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

> an install hint (`curl -fsSL https://x.ai/cli/install.sh | bash`).
commands/grok.md · 115 lines

How it starts

The opening of the file, as written. The whole thing — 115 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/cc-suite:grok

Send a prompt to Grok Build through the cc-suite runner. This is the direct Claude → Grok delegation surface, alongside the Codex- and agy-backed lanes.

Under the hood the runner (scripts/grok-runner.mjs) drives grok agent stdio as an ACP (Agent Client Protocol) client — it acts as the client, Grok is the agent. It streams the answer from session/update notifications, is deadline-bounded and killable, and registers every call as a job, so /cc-suite:status, /result, /cancel, and /continue work identically to the Codex and agy backends. Grok can call back into Claude through the claude-code MCP server if it's bridged (/cc-suite:bridge-tools with grok enabled).

The delegation boundary is injected for you. Grok reads AGENTS.md and the shared .agents/skills/ tree natively, so it can see cc-suite's own skills for delegating to Claude Code. The runner prepends lib/delegation-boundary.mjs to every prompt so a delegated task is not handed straight back to its author. Do not restate it in the prompt; it is already there.

User Input

$ARGUMENTS

Workflow

Step 1: Parse the request

Extract these optional flags from $ARGUMENTS; the remaining text is the prompt:

  • --model <id> — a Grok model id (e.g. grok-build). Omit to use Grok's configured default. Run grok models to list available ids.
  • --effort <level> — reasoning effort: none, minimal, low, medium, high, xhigh, max. Omit for the default. (Unlike agy, Grok does take an effort flag.)
  • --sandbox <read-only|workspace-write|danger-full-access> — default read-only.
  • --background or --wait — default --wait.
  • --resume <session-id> — continue a prior Grok session (the threadId a previous call returned).

If the prompt is empty, ask the user what they want Grok to do and stop if they don't provide one.

Step 2: Verify Grok is ready (fail fast)

Grok can stall until the deadline if it isn't installed or logged in, so gate on a fast local readiness check before committing to a run:

Read the full file on GitHub · 115 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.

  1. 3d ago First seen · 115 lines · 35 tokens per session scan C 7430d04b62cf

Subscribe to this mod's changes

grok is a command published in the GitHub repository xiaolai/cc-suite (45 stars, last pushed yesterday), licensed ISC. It adds 35 tokens to every session and 1,332 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.