Grok Executor

What it does

When your agent uses it

• The user explicitly asks Codex to have Grok
• Grok CLI implement
• Research something
• Ordinary tasks that do not request Grok

Example prompts

• “/grok-executor”

Requirements

• Python 3
• A Bash shell

Workflow steps

4 steps, taken from the first numbered list in SKILL.md.

1. 1Read all applicable repository instructions and inspect git status --short plus the relevant diff. Preserve unrelated user changes.
2. 2Form a concrete implementation plan. Do not wait for plan approval unless the request is ambiguous or high-impact enough to require a user…
3. 3Read references/task-contract.md and write a complete task brief. Include exact scope, constraints, acceptance criteria, validation, and…
4. 4Avoid putting sensitive values in the brief. The wrapper removes its temporary task session after Codex closes it, but task content and…

What it can do on your machine

Read from SKILL.md and the folder at commit efee6a9. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Ships 2 files in scripts/ (Python and Shell), which the agent can run.

  Shell commands in SKILL.md call:

  • git

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md. Its commands use git, which can reach the network depending on how they are called.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Grok Executor loads about 1.8k tokens when it runs, and up to ~3.3k if it reads all its reference files. Until then it costs about 103 tokens; SKILL.md has 905 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~103

When it runs · the whole SKILL.md, loaded when a task matches

~1.8k

With references · SKILL.md plus every file in references/, read only if the agent opens them

~3.3k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

• WarningTells the agent its actions are pre-authorized / not to stop for confirmationSKILL.md:22

  2. Form a concrete implementation plan. Do not wait for plan approval unless the request is ambiguous or high-impact eno

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from elliothux/open-compute at commit efee6a9, republished under its Apache-2.0 licence (© elliothux). 905 words, ~1,835 tokens.

name

grok-executor

description

Delegate a concrete, locally authorized implementation or read-only web research task from Codex to the official Grok Build CLI. Use when the user explicitly asks Codex to have Grok, SuperGrok, or Grok CLI implement, inspect, test, or research something. Do not trigger for ordinary tasks that do not request Grok, or for external commits, pushes, deployments, publication, or production mutations.

.codex/skills/grok-executor/scripts/run-grok-executor.sh \
  --execute \
  --cwd /absolute/path/to/repo \
  --prompt-file /absolute/path/to/task-brief.md

.codex/skills/grok-executor/scripts/run-grok-executor.sh \
  --inspect \
  --cwd /absolute/path/to/repo \
  --prompt-file /absolute/path/to/task-brief.md

{"type":"interject","text":"Stop the single-crate approach; keep the requested workspace boundaries."}
{"type":"prompt","text":"The clippy run failed with this exact error. Fix it and rerun the scoped checks."}
{"type":"prompt_file","path":"/absolute/path/to/follow-up.md"}
{"type":"status"}
{"type":"cancel"}
{"type":"close"}

© elliothux, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 7 other files (scripts, references) in .agents/skills/grok-executor of elliothux/open-compute.

• SKILL.md
• agents/openai.yaml
• references/grok-build.md
• references/task-contract.md
• scripts/grok-acp-executor.py
• scripts/run-grok-executor.sh
• tests/fake-grok-acp.py
• tests/test_grok_acp_executor.py

Open the folder on GitHubat commit efee6a9