tubecreate/tubecli/codex
Claude Code and agent skill by tubecreate/tubecli
- community
Checking that the SKILL.md file is still in the repository.
Mission control: create tasks, assign to an agent/team, wait for human approval, worker runs in the background, accept results.
Install
npx vskill@latest install tubecreate/tubecli/codexSKILL.md
π Codex β work control panel
Codex is where ALL work that truly needs doing lives. Any request that runs long, is resource-heavy (LLM, browser, download/upload), changes data or the outside world, or needs someone to review the result β do NOT do it inline; create a Codex task and report the task number to the user. Only answer directly when the question is a quick lookup/explanation that produces no change.
π Approval rules (MANDATORY)
- Tasks created by the AI are ALWAYS in
pending_approvalstate and wait for human approval. The AI must not approve its own tasks. - After creating a task, NEVER say the work has run/finished. Only say: created
#<n>, awaiting approval, and instruct the user to typeapprove <n>. - Lifecycle:
pending_approval β queued β running β review β done(+failedcan be retried,rejected,cancelled).
β οΈ FLAT PAYLOAD rule (MANDATORY)
The action parser can only read JSON that is one level deep, all scalar values (string/number/bool). Do NOT use nested objects, do NOT use arrays in an action. To split a large job into smaller ones β emit MULTIPLE consecutive `codex_create_task` actions, one goal per action. Never pack a list of subtasks into a single action.
π₯ codex_create_task β create a new task
goal(required) β a self-contained description; the worker reads only this string.assignee_type:"agent"or"team".assigneetakes a NAME (matching an existing agent/team name), no ID needed.priority: the higher the number, the sooner it runs (default 0).
π₯ codex_list_tasks β view the list
status: active (default) | pending_approval | queued | running | review | done | failed | rejected | cancelled.
π₯ codex_task_status β details of one task
task takes a short number (3), the full id, or part of the title.
π₯ codex_approve / codex_reject β decide on the user's behalf
Only use when the user explicitly states they want to approve/reject.
π₯ codex_cancel β cancel a pending/running task
π₯ codex_retry β re-run a failed or review task
π HTTP API (via run_api, prefix /api/v1/codex)
Prefer the codex_* actions above. Only use run_api for GET endpoints not in the action list.
| Method | Endpoint | Purpose | |
|---|---|---|---|
| GET | /api/v1/codex/stats | Count tasks by status | |
| GET | /api/v1/codex/tasks?status=&limit= | Task list | |
| POST | /api/v1/codex/tasks | Create a task (use codex_create_task instead) | |
| GET | /api/v1/codex/tasks/{id} | Task + event log | |
| GET | /api/v1/codex/tasks/{id}/events?after=&limit= | Event log | |
| POST | /api/v1/codex/tasks/{id}/approve Β· /reject Β· /cancel Β· /retry | Decisions | |
| POST | /api/v1/codex/tasks/{id}/review | Acceptance (accepted: true/false) | |
| POST | /api/v1/codex/tasks/{id}/plan | AI breaks down the goal (SLOW 10β60s) | |
| GET | /api/v1/codex/assignees | List of agents + teams | |
| GET | /api/v1/codex/worker | Worker status |
Visual dashboard: GET /codex.
β¨οΈ Text commands (0 token β the user types them directly, the AI need not process them)
codex Β· codex <n> Β· approve <n> Β· reject <n> Β· retry <n> Β· accept <n> Β· codex cancel <n> Β· codex running|done|failed (A bare Β«cancelΒ»/Β«huα»·Β» is already claimed by the bot's plan-confirmation flow, so cancelling a task must be typed with the codex prefix.) When the user's message matches this form, do not emit an action β the system handles it itself.
π‘ Examples
User: "Research 5 of my competitor YouTube channels, then write a report" β long, resource-heavy, needs review β create a task, don't do it yourself:
User: "Translate this article then post it to WordPress, and once done create an illustration" β two jobs β two separate actions, don't cram them into an array:
User: "Are there any tasks waiting for my approval?"
User: "How is task 3 going?"
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingtubecli/extensions/codex/SKILL.md
Popularity
- 7-day trend score
- 0
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/tubecreate/tubecli/codex)