🗺️ Data Science & Analytics · Journey Playbooks

Playbook: Produce an analysis stakeholders trust

From raw tables to defensible findings with documented caveats.

journey~85 minData ScientistsAnalystsResearch Scientists

Journey Steps

1Step 1 — Run a dataset health check before modeling: start with "Check shape, dtypes, and unique counts; flag columns failing expectations"
2Step 2 — Avoid correlation traps in stakeholder analyses: start with "Ask what could drive both variables; list top three confounders"
3Step 3 — Write metric definitions as contracts: start with "Create a metric catalog entry per KPI: formula, grain, filters, owner"
4How it fits together: Health check before insight. One canonical metric definition prevents the follow-up war.

Commands

Install all referenced skills

$ npx skills add aniruddhaadak80/skills --skill exploratory-analysis-dataset-health-check && npx skills add aniruddhaadak80/skills --skill exploratory-analysis-correlation-causation && npx skills add aniruddhaadak80/skills --skill experimentation-metric-definition-contracts

Tags

#playbook#journey#data-science

Install this skill

$ npx skills add aniruddhaadak80/skills --skill playbook-data-science-analysis-trust -a opencode

SKILL.md

# Playbook: Produce an analysis stakeholders trust

> From raw tables to defensible findings with documented caveats.

## When to Use This Skill

From raw tables to defensible findings with documented caveats.

## Steps

1. Step 1 — Run a dataset health check before modeling: start with "Check shape, dtypes, and unique counts; flag columns failing expectations"
2. Step 2 — Avoid correlation traps in stakeholder analyses: start with "Ask what could drive both variables; list top three confounders"
3. Step 3 — Write metric definitions as contracts: start with "Create a metric catalog entry per KPI: formula, grain, filters, owner"
4. How it fits together: Health check before insight. One canonical metric definition prevents the follow-up war.

View source on GitHub ↗

Take retrieval-augmented answers from empty repo to evaluated production feature.

🗺️ AI Engineering·~135m

Open →

🗺️Playbook: Harden an agent before giving it toolsjourney

Ship an autonomous agent whose failure modes are cheap, visible, and reversible.

🗺️ AI Engineering·~125m

Open →

🗺️Playbook: Cut LLM costs without quality collapsejourney

Reduce monthly inference spend measurably while keeping answer quality within tolerance.

🗺️ AI Engineering·~95m

Open →