Guide · Agent skill
How to use the cheat-status agent skill: content project dashboard
Updated · written with AI from the project's own documentation and checked against our catalogue data
cheat-status is an agent skill for the cheat-on-content short-video workflow. When you ask your agent for the status, it reads your project's state file and prediction, candidate and rubric files, then returns a dashboard of what to post, review or prepare next. This guide explains what the skill does, how to install it, how to phrase requests so it triggers, and what it needs to run.
Content Project Status DashboardXBuilderLAB/cheat-on-content · MIT Download the skill Quick start
- 1Download the ZIP. Get the skills/cheat-status ZIP from the directory page.
- 2Unzip into a skills folder. Extract it to ~/.claude/skills/ for personal use, or to a project's .claude/skills/ folder.
- 3Or upload in the Claude apps. In the Claude apps, open Settings → Capabilities → Skills and upload the ZIP.
- 4Open your cheat-on-content project. Start the agent from inside the project folder that contains .cheat-state.json.
- 5Ask for the status. Type a trigger such as 状态, 看板, status or 我现在该做什么.
- 6Act on the top suggestion. Run the first command in the prioritised list, then ask for the status again.
What is an agent skill, and what does cheat-status do?
An agent skill is a folder containing a SKILL.md file that tells an AI agent how to perform a specific task. The agent loads the skill when a request matches it. The source for cheat-status describes a skill that is designed to be read-only: it reads files and reports, but does not change them.
The skill reads a local state file (.cheat-state.json) and scans your project's predictions/ folder, candidates.md and rubric_notes.md. From these it summarises the current mode, rubric version, calibration progress, videos waiting for review, the upcoming-video buffer, and whether to upgrade storage or bump the rubric.
- Current content form, target cadence and rubric version
- Calibration sample count and confidence level
- Buffer size with a colour indicator based on days of buffer
- Videos due for review (retro)
- Candidate pool size by tier
- Warnings for stale in-progress predictions
- A prioritised list of concrete commands to run next
Install the cheat-status skill
Download the ZIP for skills/cheat-status from the directory page. The skill is distributed as a folder containing SKILL.md. There are two ways to install it.
Option 1, personal install: unzip the file into ~/.claude/skills/ so the cheat-status folder sits inside that directory. Option 2, project install: unzip it into the .claude/skills/ folder of a specific project.
Option 3, Claude apps: upload the ZIP in the Claude apps under Settings → Capabilities → Skills.
- Personal: ~/.claude/skills/cheat-status/
- Project: <your-project>/.claude/skills/cheat-status/
- Claude apps: Settings → Capabilities → Skills, then upload the ZIP
Check your project is set up
The skill only shows useful information when a cheat-on-content project exists. The project folder needs .cheat-state.json, a predictions/ folder, candidates.md and rubric_notes.md. The skill's allowed tools are Bash, Read, Glob and Grep.
If .cheat-state.json is missing, the skill does not produce a dashboard. It tells you that you have not initialised yet and to run /cheat-init first. Run the agent from inside your cheat-on-content project folder so the relative file paths resolve.
- Required: .cheat-state.json, predictions/, candidates.md, rubric_notes.md
- Required tool access: Bash, Read, Glob, Grep
- Not set up yet? The skill tells you to run /cheat-init
How to trigger the skill with natural requests
The skill's description lists these trigger phrases: 状态, 看板, status, 我现在该做什么 (what should I do now) and 进度怎么样 (how is progress). You can also use the slash form /cheat-status.
Ask in the same plain way you would ask a colleague. The skill is meant to be called at any time, so you can run it before posting, after a review, or whenever you want a checklist.
- 状态
- 看板
- status
- 我现在该做什么
- 进度怎么样
- /cheat-status
What the dashboard looks like and how to act on it
The output is a dashboard that starts with the content form, rubric version and calibration status, then lists the buffer, progress bars, a to-do list ordered by urgency, pool status and a health section. Each suggestion comes with an exact command you can copy and run.
The skill ranks suggestions by urgency so the most important item appears first. Acting on a suggestion means running the command it names, such as a retro or a bump. The status skill itself does not run those actions.
- Red buffer: post at least one video today, per the skill's guidance
- Blue buffer: pause filming and publish stored videos first
- Videos past the review window are listed for retro
- Health section shows rubric_notes.md line count and whether hooks are installed
Safety notes: scripts, network and file changes
The skill's metadata says it runs scripts and does not use the network. It is declared with Bash(*) access, so it can run shell commands in your project. The source describes it as read-only, meaning it reads and reports but does not modify state files, predictions or candidates. Changes are left to other skills.
Because it can run Bash, review the SKILL.md before installing and run it inside a project folder you trust. The source states it makes no network calls.
- Runs scripts: yes (Bash access)
- Uses network: no
- Read-only by design; state changes belong to other skills
- Review SKILL.md before installing if you want to inspect what commands it runs
What people use it for
- Daily checklist: see whether the buffer is healthy and what to film or publish today
- Review planning: find predictions whose review window has passed
- Rubric timing: check whether calibration samples suggest a rubric bump
- Stale work check: spot in-progress predictions that have not been published for a long time
- Pool check: see how many candidate topics are in the pool and whether a refresh is due
Frequently asked questions
How do I install the cheat-status skill?
Download the ZIP for skills/cheat-status from the directory page and unzip it into ~/.claude/skills/ for personal use or a project's .claude/skills/ folder. You can also upload the ZIP in the Claude apps under Settings → Capabilities → Skills.
Is cheat-status free and what is its licence?
The source lists the licence as MIT. The source does not state pricing, so check the directory page for any cost information.
What do I need to use cheat-status?
You need a cheat-on-content project folder containing .cheat-state.json, predictions/, candidates.md and rubric_notes.md, plus Bash, Read, Glob and Grep tool access in your agent. Without the state file, the skill only tells you to run init.
Does cheat-status change my files?
No. The skill is designed to be read-only and has no side effects. It reads state and project files and reports, while actions such as reviews or updates are handled by other skills.
Why does cheat-status say I have not initialised?
It only produces a dashboard when .cheat-state.json exists. If the file is missing, it tells you to run /cheat-init first. Make sure you start the agent from inside your project folder.
Does cheat-status use the internet or run scripts?
The metadata says it runs scripts and does not use the network. It has Bash access, so review SKILL.md before installing if you want to see the exact commands.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
cheat-status suits creators already using the cheat-on-content workflow who want a daily, read-only checklist of what to film, review or prepare. It is not useful without that project set up, and it will not publish, review or edit anything itself. If you need a general project tracker rather than this specific workflow, pick a tool built for that purpose.
Download the skill