Guide · Agent skill
How to use the CDO Plan Review agent skill (cdo-review)
Updated · written with AI from the project's own documentation and checked against our catalogue data
The cdo-review skill is a checklist and report template that makes your AI agent review a data plan the way a Chief Data Officer would. It is useful for founders deciding whether to train models on customer data, data leaders choosing between warehouse, lakehouse or mesh architectures, and teams preparing data assets for licensing or M&A. After this guide you will know what the skill does, how to install it, how to trigger it with plain requests, and what it does not cover.
CDO Plan Reviewalirezarezvani/claude-skills · MIT Download the skill Quick start
- 1Download the ZIP. Get the cdo-review ZIP from this page. It contains SKILL.md.
- 2Unzip into a skills folder. Extract it to ~/.claude/skills/ for personal use, or to a project's .claude/skills/ folder to limit it to one project.
- 3Install the parent plugin. Install the claude-skills c-level-advisor plugin so the vendor paths referenced by the skill resolve.
- 4Make a request with your plan. Ask in plain language about training data, architecture, data productization or a data hire, and include the plan details.
- 5Review the verdict. Read the SHIP, SHARPEN or BLOCK verdict and the sections that apply to your plan.
- 6Act on next steps. Work through the three next steps and consult a lawyer for any consent or contract conclusions.
What an agent skill is and what cdo-review makes the agent do
An agent skill is a folder containing a SKILL.md file that gives an AI agent instructions for a specific kind of task. When a request matches the skill's description, the agent follows its checklist and output format.
When you ask the agent to review a plan that touches data strategy, cdo-review runs the plan through six Chief Data Officer-style questions: what decision the data supports, where each source's consent comes from, how many teams consume it, whether it would survive an acquirer's diligence, whether the work could proceed without that source, and which hire is actually needed.
The skill then produces a structured report with a SHIP, SHARPEN or BLOCK verdict and next steps. The report includes sections for training audit, architecture, asset value and org, which are filled in only where they apply.
Install the cdo-review skill from the ZIP
Download the ZIP for cdo-review from this page and unzip it. The skill is delivered as a single file, SKILL.md.
For a personal install, unzip it into ~/.claude/skills/. To limit the skill to one project, unzip it into that project's .claude/skills/ folder.
You can also upload the skill in the Claude apps under Settings → Capabilities → Skills.
- Personal install: ~/.claude/skills/
- Project install: .claude/skills/ inside your project
- Claude apps: Settings → Capabilities → Skills (upload)
Install the parent plugin so referenced paths resolve
The skill refers to vendor paths under the c-level-advisor plugin, such as the chief-data-officer-advisor skill and its scripts. Those paths only resolve if the parent claude-skills c-level-advisor plugin is installed.
Without it, the skill's text checklist still works, but the references to companion files and scripts will point nowhere.
Trigger the skill with natural requests
The skill's description names its trigger situations: validating training-data rights before model work, choosing warehouse vs lakehouse vs mesh, and valuing data assets for productization or M&A. Ask for one of these in plain language and include the plan itself.
The source also documents a slash-command form, /c-level-agents:cdo-review [plan], followed by your plan. Use whichever form your setup supports.
- Before we train the support-ticket model on customer chats, review the consent situation and give me a verdict.
- Review this plan: we are choosing between a warehouse and a lakehouse for 12 analytics teams. Use the CDO questions.
- We want to license our anonymized usage data to a partner. Run a CDO review before we sign anything.
/c-level-agents:cdo-review <plan>Read the output and the next steps
The report starts with the decision being made, which the template classifies as training, architecture, asset or hire. It then fills in only the sections that apply to your plan.
The verdict is one of 🟢 SHIP, 🟡 SHARPEN or 🔴 BLOCK. The report ends with three concrete next steps. The routing section of the skill suggests other reviews to run, such as a legal review for productization or a CISO review for architecture changes touching customer data.
- Training audit: counts of NO-GO, MITIGATE and GO sources, plus a top remediation
- Architecture: recommended WAREHOUSE, LAKEHOUSE or MESH, with kill criteria
- Asset value and org sections, when relevant
- Verdict and three next steps
Know what the companion scripts need and what they do not do
The skill mentions three Python scripts for the training-data audit, architecture choice and asset valuation. They are not included in the reviewed files, so the skill cannot run them as delivered.
Without the scripts, the report's numeric fields such as valuation multiples have no reliable source. Treat those figures as unverified and do not use them for M&A decisions.
The skill runs no scripts itself and makes no network calls, according to the reviewed file.
- Requires the parent claude-skills c-level-advisor plugin for referenced paths
- Python is needed only if you run the optional companion scripts
- Runs scripts: no. Uses network: no.
What people use it for
- Deciding whether a support-ticket or chat dataset can be used to train a model, with a verdict on consent sources.
- Choosing between warehouse, lakehouse or mesh when a growing number of analytics teams need the same data.
- Checking whether a customer dataset is ready for a licensing deal or an acquirer's diligence questions.
- Reviewing a hire for a head of data or ML engineer to confirm prerequisite roles are in place.
- Pressure-testing a multi-year data infrastructure contract before signing.
Frequently asked questions
Is the cdo-review skill free to use?
The source lists the skill under the MIT licence. The reviewed files do not mention pricing for any other part, so check the repository for any additional terms.
What do I need to run cdo-review?
You need the parent claude-skills c-level-advisor plugin installed so referenced paths resolve. Python is only needed if you want to run the optional companion scripts, which are not included in the reviewed file.
How do I install a Claude agent skill?
Unzip the downloaded ZIP into ~/.claude/skills/ for personal use or into a project's .claude/skills/ folder. You can also upload it in the Claude apps under Settings → Capabilities → Skills.
Does cdo-review give legal advice on training data consent?
No. Its consent and contract conclusions are general prompts, not a substitute for a lawyer reviewing actual terms. Regulatory exposure is high for personal data, so involve counsel before acting.
Can cdo-review give me an accurate M&A valuation?
Not from the skill alone. The output template expects numbers from scripts that are not in the reviewed files, so any figures the agent produces would be unreliable. Use a proper valuation process for deal math.
Is cdo-review worth it for a small project?
Probably not if you have no customer data and no architecture decision. Six formal questions add overhead to small projects with nothing to decide about data.
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
Use cdo-review if you are a founder or data leader who needs a structured, decision-focused challenge before training on customer data, picking an architecture, or preparing data for licensing. Pick something else if you need legal opinions, precise valuation numbers, or a lightweight review for a small project with no customer data.
Download the skill