grok-prompting

grok-prompting is a skill for Claude Code from zachdunn/grok-plugin-claude-code. It costs 73 tokens per session (534 once invoked), scanned A, original, MIT.

A guide for writing precise prompts before sending a task to Grok, an AI model that runs the task in one headless pass without back-and-forth questions.

In plain words
What is it for?
Preparing Grok delegation prompts that specify what to change or report, where to work, what to leave untouched, and how completion will be checked.
Why use it?
It turns vague requests into tasks with a clear goal, location, constraints, and success conditions.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: mentions subagents.

Part of the grok-cc plugin — 3 skills, 4 commands, 1 agent, 1 hook shipped together

Good fit Preparing Grok delegation prompts that specify what to change or report, where to work, what to leave untouched, and how completion will be checked.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/zachdunn/grok-plugin-claude-code/grok-prompting
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.

Any agent
npx skills add zachdunn/grok-plugin-claude-code --skill grok-prompting
Clone the repo
git clone --depth 1 https://github.com/zachdunn/grok-plugin-claude-code

Made for: Claude Code.

Or install grok-cc, the plugin that ships this one along with the rest of its 3 skills, 4 commands, 1 agent, 1 hook.

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-prompting

README.md
[![agentmods](https://agentmods.dev/badge/skills/zachdunn/grok-plugin-claude-code/grok-prompting/github.svg)](https://agentmods.dev/skills/zachdunn/grok-plugin-claude-code/grok-prompting)
Your own site
<a href="https://agentmods.dev/skills/zachdunn/grok-plugin-claude-code/grok-prompting"><img src="https://agentmods.dev/badge/skills/zachdunn/grok-plugin-claude-code/grok-prompting/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.

agentmods 80×15 button for grok-prompting

Your own site · 80×15
<a href="https://agentmods.dev/skills/zachdunn/grok-plugin-claude-code/grok-prompting"><img src="https://agentmods.dev/badge/skills/zachdunn/grok-plugin-claude-code/grok-prompting.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 73 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 534 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00073 $0.00534
Opus 5 $0.00036 $0.00267
Sonnet 5 $0.00015 $0.00107
Haiku 4.5 $0.00007 $0.00053

Measured 9d ago against content hash e3c6f4a0fc53, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

grok-prompting 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 9d 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.

plugins/grok-cc/skills/grok-prompting/SKILL.md · 43 lines

What it actually says

Grok prompting

Grok runs the delegated task in a single headless pass — there is no back-and-forth to clarify. The prompt has to carry everything Grok needs. Shape the user's request into a crisp task before forwarding it; do not do the task yourself.

What a good Grok task prompt contains

  1. The goal, in one sentence. What should be true when Grok is done.
  2. Where to work. Name the directory, package, or files when known. Grok starts from the run's cwd and reads from there.
  3. Constraints. What it may and may not touch; conventions to follow; anything to leave alone.
  4. Acceptance criteria. How Grok (and the user) will know it worked — "tests pass", "tsc clean", "the endpoint returns 200", a specific behavior.
  5. Done-ness. For a write task, ask for the edits plus a short summary of what changed. For a read task, ask for the specific answer/finding format.

Keep it tight

  • Prefer concrete nouns (file paths, function names, error text) over vague description. Paste the actual error, not a paraphrase.
  • State the outcome, not a step-by-step procedure — let Grok plan the steps.
  • One task per run. Bundle only tightly-related changes.
  • Pick the mode deliberately: write-capable (default) to make changes, --read for review/diagnosis/research with no edits.
  • Reach for --effort high on genuinely hard problems; leave it default otherwise. Use --check when correctness matters more than latency.

Anti-patterns

  • Don't forward a one-word ask ("fix it") — add the goal and acceptance check.
  • Don't include routing flags (--read, --effort, …) inside the task text; they are controls, not instructions for Grok.
  • Don't ask Grok to do exploratory work the main thread should just do itself — delegate substantial, bounded tasks, not trivial lookups.
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. 9d ago First seen · 43 lines · 73 tokens per session scan A e3c6f4a0fc53

Subscribe to this mod's changes

grok-prompting is a skill published in the GitHub repository zachdunn/grok-plugin-claude-code (30 stars, last pushed 2mo ago), licensed MIT. It adds 73 tokens to every session and 534 once invoked, about $0.0004 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.