Grok Build Orchestration
When to Use
- Use when delegating a well-specified implementation task to xAI's Grok Build CLI running headlessly
- Use when executing a Markdown implementation plan task-by-task with a diff review after each task
- Use when the user says "use grok", "grok build", "have grok implement", or "send to grok"
The coding assistant is the orchestrator: it plans, writes self-contained task specs,
dispatches them to Grok Build headlessly, reviews every diff, and owns the final result.
Grok is the fast, cheap executor. Full CLI details and verified behaviors: references/cli.md.
Safety Gate
Before every dispatch, show the user the exact task specification that will be sent to xAI,
the target worktree, and the permission mode. Obtain explicit approval to disclose that text
and to let Grok edit the scoped worktree. Never include secrets, proprietary source, customer
data, or credentials in a task specification. Do not run grok update, --always-approve,
or a destructive recovery command without separate, explicit approval.
When to delegate vs keep with the orchestrator
| Delegate to Grok |
Keep with the orchestrator |
| Plan tasks with clear acceptance criteria |
Ambiguous requirements, architecture decisions |
| Boilerplate, scaffolding, CRUD |
Deep cross-file debugging |
| Mechanical refactors |
Security-sensitive code |
| Test writing from clear specs |
Anything touching production infrastructure |
| UI components from mockups/specs |
Tasks where writing the spec ≈ doing the work |
When in doubt, keep it with the orchestrator.
Session preflight (once, before the first dispatch)
grok update --check --json — if updateAvailable is true, tell the user. Run
grok update only after explicit approval, then confirm with grok --version.
grok models — if it errors or reports logged out, STOP and ask the user to run
grok login.
Per-task loop (sequential — the default)
Spec. Write a self-contained task file (template below) to a temp directory
OUTSIDE the target repo — the harness scratchpad if one is available, else the OS
temp dir. Never write it inside the target repo. Grok has zero conversation context:
no one-liner prompts, ever.
- POSIX:
mkdir -p "${TMPDIR:-/tmp}/grok-specs", then write task.md there.
- Windows (PowerShell):
New-Item -ItemType Directory -Force "$env:TEMP\grok-specs",
then write task.md there.
Clean state. No uncommitted source changes — commit or stash first, so the
post-run diff is exactly Grok's work. Ignore build artifacts (__pycache__, dist/,
etc.); if they show in git status, they're usually just un-gitignored, not your
concern. Never dispatch on a dirty source tree.
Dispatch.
POSIX:
grok --prompt-file <task-file> \
--output-format json \
--always-approve \
--max-turns 30 \
--cwd <repo>
Windows (PowerShell) — backtick line-continuation:
grok --prompt-file <task-file> `
--output-format json `
--always-approve `
--max-turns 30 `
--cwd <repo>
Parse the JSON output and save sessionId. (--always-approve is required for
headless runs — --permission-mode acceptEdits silently cancels edits with no
interactive approver. Use it only after the user explicitly approves Grok editing this
exact scoped worktree. See references/cli.md.) For a high-stakes task, add --check
so Grok self-verifies before you review; skip it otherwise (it ~doubles latency).
Review gate — non-negotiable.
- Read the diff yourself (
git diff -- <files from the spec> to skip artifact noise):
does it do the task, only the task, and match repo conventions?
- Run the acceptance commands from the spec.
- Pass → commit with a clear message following the repo's convention → next task.
- Fail → ask the user before a fix-up or any reset. Never run
git checkout -- . or
git clean -fd automatically; preserve the diff for review and use a non-destructive
recovery plan unless the user explicitly authorizes otherwise.
Task spec template
# Task: <one-line title>
## Context
- Repo: <path> — <one line on what the project is>
- Conventions: <test runner, formatter, a good example file to imitate>
## Files
- Modify: <path>
- Create: <path>
## Task
<precise description of the change>
## Constraints
- Do not modify any files other than those listed above.
- <other constraints>
## Acceptance criteria
- `<exact command>` <expected result>
Executing a Markdown implementation plan
- One plan task per dispatch, in order.
- Check off the plan's task checkboxes (
- [ ] → - [x]) as each task lands and passes
the review gate.
- If the plan explicitly marks tasks as independent, see Parallel dispatch below;
otherwise stay sequential.
Parallel dispatch (opt-in exception, not the default)
Only when a plan explicitly marks tasks independent: dispatch each with
--worktree=<task-slug>, run concurrently, then review and merge one worktree at a
time through the same review gate. Merge conflicts usually eat the savings — prefer
sequential.
Failure handling
| Failure |
Action |
stopReason: "Cancelled", empty text, no diff |
Missing --always-approve — retry with it |
| CLI error / timeout |
Retry once; then do the task yourself and note the fallback |
| Auth expired |
Stop; ask the user to run grok login |
| 2 fix-up rounds exhausted |
Preserve the diff, ask the user for a recovery decision, then finish the task manually if authorized |
| Dirty tree at dispatch |
Refuse; commit/stash first |
Limitations
- Grok receives the approved task specification; it is a third-party service and should not
receive secrets, proprietary material, personal data, or customer data.
--always-approve allows edits without an interactive approval prompt. It must be limited to
a clean, explicitly approved worktree and never substitutes for the orchestrator's review.
- Model output can be incorrect, insecure, incomplete, or out of scope. Review the diff and
run the acceptance checks before accepting any change.
- This skill does not authorize installations, updates, commits, pushes, deployments, or
destructive cleanup.
Models
Default grok-4.5. Add -m grok-composer-2.5-fast only for trivial mechanical tasks.
1---2name: grok-build3description: "Delegate well-specified implementation tasks to xAI's Grok Build CLI running headlessly while the orchestrating agent plans, writes task specs, reviews every diff, and owns the result."4category: agent-orchestration5risk: critical6source: https://github.com/sanjay3290/ai-skills/tree/main/skills/grok-build7source_repo: sanjay3290/ai-skills8source_type: community9date_added: "2026-07-09"10author: sanjay329011tags: [grok, delegation, code-generation, xai]12tools: [claude, cursor, gemini]13license: "Apache-2.0"14license_source: "https://github.com/sanjay3290/ai-skills/blob/main/LICENSE"15metadata:16 source: https://skillmd.com/skills/sanjay3290/grok-build17---1819# Grok Build Orchestration2021## When to Use2223- Use when delegating a well-specified implementation task to xAI's Grok Build CLI running headlessly24- Use when executing a Markdown implementation plan task-by-task with a diff review after each task25- Use when the user says "use grok", "grok build", "have grok implement", or "send to grok"2627The coding assistant is the orchestrator: it plans, writes self-contained task specs,28dispatches them to Grok Build headlessly, reviews every diff, and owns the final result.29Grok is the fast, cheap executor. Full CLI details and verified behaviors: `references/cli.md`.3031## Safety Gate3233Before every dispatch, show the user the exact task specification that will be sent to xAI,34the target worktree, and the permission mode. Obtain explicit approval to disclose that text35and to let Grok edit the scoped worktree. Never include secrets, proprietary source, customer36data, or credentials in a task specification. Do not run `grok update`, `--always-approve`,37or a destructive recovery command without separate, explicit approval.3839## When to delegate vs keep with the orchestrator4041| Delegate to Grok | Keep with the orchestrator |42|---|---|43| Plan tasks with clear acceptance criteria | Ambiguous requirements, architecture decisions |44| Boilerplate, scaffolding, CRUD | Deep cross-file debugging |45| Mechanical refactors | Security-sensitive code |46| Test writing from clear specs | Anything touching production infrastructure |47| UI components from mockups/specs | Tasks where writing the spec ≈ doing the work |4849When in doubt, keep it with the orchestrator.5051## Session preflight (once, before the first dispatch)52531. `grok update --check --json` — if `updateAvailable` is true, tell the user. Run54 `grok update` only after explicit approval, then confirm with `grok --version`.552. `grok models` — if it errors or reports logged out, STOP and ask the user to run56 `grok login`.5758## Per-task loop (sequential — the default)59601. **Spec.** Write a self-contained task file (template below) to a temp directory61 OUTSIDE the target repo — the harness scratchpad if one is available, else the OS62 temp dir. Never write it inside the target repo. Grok has zero conversation context:63 no one-liner prompts, ever.64 - POSIX: `mkdir -p "${TMPDIR:-/tmp}/grok-specs"`, then write `task.md` there.65 - Windows (PowerShell): `New-Item -ItemType Directory -Force "$env:TEMP\grok-specs"`,66 then write `task.md` there.672. **Clean state.** No uncommitted *source* changes — commit or stash first, so the68 post-run diff is exactly Grok's work. Ignore build artifacts (`__pycache__`, `dist/`,69 etc.); if they show in `git status`, they're usually just un-gitignored, not your70 concern. Never dispatch on a dirty source tree.713. **Dispatch.**7273 POSIX:7475 ```bash76 grok --prompt-file <task-file> \77 --output-format json \78 --always-approve \79 --max-turns 30 \80 --cwd <repo>81 ```8283 Windows (PowerShell) — backtick line-continuation:8485 ```powershell86 grok --prompt-file <task-file> `87 --output-format json `88 --always-approve `89 --max-turns 30 `90 --cwd <repo>91 ```9293 Parse the JSON output and save `sessionId`. (`--always-approve` is required for94 headless runs — `--permission-mode acceptEdits` silently cancels edits with no95 interactive approver. Use it only after the user explicitly approves Grok editing this96 exact scoped worktree. See `references/cli.md`.) For a high-stakes task, add `--check`97 so Grok self-verifies before you review; skip it otherwise (it ~doubles latency).984. **Review gate — non-negotiable.**99 - Read the diff yourself (`git diff -- <files from the spec>` to skip artifact noise):100 does it do the task, only the task, and match repo conventions?101 - Run the acceptance commands from the spec.102 - **Pass** → commit with a clear message following the repo's convention → next task.103 - **Fail** → ask the user before a fix-up or any reset. Never run `git checkout -- .` or104 `git clean -fd` automatically; preserve the diff for review and use a non-destructive105 recovery plan unless the user explicitly authorizes otherwise.106107## Task spec template108109```markdown110# Task: <one-line title>111112## Context113- Repo: <path> — <one line on what the project is>114- Conventions: <test runner, formatter, a good example file to imitate>115116## Files117- Modify: <path>118- Create: <path>119120## Task121<precise description of the change>122123## Constraints124- Do not modify any files other than those listed above.125- <other constraints>126127## Acceptance criteria128- `<exact command>` <expected result>129```130131## Executing a Markdown implementation plan132133- One plan task per dispatch, in order.134- Check off the plan's task checkboxes (`- [ ]` → `- [x]`) as each task lands and passes135 the review gate.136- If the plan explicitly marks tasks as independent, see Parallel dispatch below;137 otherwise stay sequential.138139## Parallel dispatch (opt-in exception, not the default)140141Only when a plan explicitly marks tasks independent: dispatch each with142`--worktree=<task-slug>`, run concurrently, then review and merge one worktree at a143time through the same review gate. Merge conflicts usually eat the savings — prefer144sequential.145146## Failure handling147148| Failure | Action |149|---|---|150| `stopReason: "Cancelled"`, empty text, no diff | Missing `--always-approve` — retry with it |151| CLI error / timeout | Retry once; then do the task yourself and note the fallback |152| Auth expired | Stop; ask the user to run `grok login` |153| 2 fix-up rounds exhausted | Preserve the diff, ask the user for a recovery decision, then finish the task manually if authorized |154| Dirty tree at dispatch | Refuse; commit/stash first |155156## Limitations157158- Grok receives the approved task specification; it is a third-party service and should not159 receive secrets, proprietary material, personal data, or customer data.160- `--always-approve` allows edits without an interactive approval prompt. It must be limited to161 a clean, explicitly approved worktree and never substitutes for the orchestrator's review.162- Model output can be incorrect, insecure, incomplete, or out of scope. Review the diff and163 run the acceptance checks before accepting any change.164- This skill does not authorize installations, updates, commits, pushes, deployments, or165 destructive cleanup.166167## Models168169Default `grok-4.5`. Add `-m grok-composer-2.5-fast` only for trivial mechanical tasks.