CDO Forcing Questions
The CDO Forcing Questions skill pressure-tests any plan touching data strategy through six decision-driven questions covering training-data rights, architecture choice, data productization, and data team hiring. It produces a structured CDO review report with a verdict and next steps.
✨ What it does
- Checks consent provenance for every data source
- Determines architecture recommendation (warehouse, lakehouse, mesh) based on consumer count
- Assesses M&A diligence impact and data asset valuation
- Tests dependency on specific data sources via retrainability
- Maps the decision to the correct next hire role
- Generates a verdict (SHIP, SHARPEN, BLOCK) with next steps
🎯 When to use it
- Before approving a new ML model training run that uses customer data
- Before signing a multi-year data-infrastructure SaaS contract
- Before productizing customer data or entering M&A diligence
- Before a major data team hire
- When the founder uses the word 'monetize' near 'data'
🚀 How to use
Trigger: /cs:cdo-review
Trigger with the slash command /cs:cdo-review <plan> where <plan> is the plan text or file. The skill expects a description of the plan that touches data strategy. Example prompts:
/cs:cdo-review We plan to train a foundation model on customer support transcripts.
/cs:cdo-review We are evaluating Snowflake vs Databricks for our new data platform.
📄 Output: A structured CDO review report in markdown with sections for decision, training audit, architecture, asset value, org, verdict, and next steps.
📦 Add this skill to Claude Code
# 1. Get the skills repo
git clone --depth 1 https://github.com/alirezarezvani/claude-skills /tmp/claude-skills
# 2. Copy this skill into your project (or ~/.claude/skills for all projects)
mkdir -p .claude/skills
cp -r /tmp/claude-skills/c-level-agents/skills/cdo-review .claude/skills/cdo-reviewSkill source: c-level-agents/skills/cdo-review/SKILL.md
⚠️ Good to know
Requires access to the referenced scripts and profile/source files for full automation; otherwise manual input is needed.
❓ FAQ
What are the six CDO questions?
The six questions are: 1) What decision does this data drive? 2) What's the consent provenance for every source? 3) Who consumes this internally and how many domains? 4) What's the M&A diligence impact? 5) Can the model/decision/report be retrained without this source? 6) What role unblocks this and is it the right hire?
How does the skill decide between warehouse, lakehouse, and mesh?
It uses consumer count: fewer than 5 consumers suggests warehouse-only, 5-25 suggests lakehouse, and 25+ with federated culture suggests mesh.
What verdicts can the skill output?
The verdict can be SHIP (green), SHARPEN (yellow), or BLOCK (red).
🤖 Overview, features, install steps and FAQ were generated from the project's SKILL.md on Sep 4, 2026. Always check the original source before running commands.