# Case Study Engine

> Discover evidence-backed case-study candidates and draft collateral for human approval without inferring causation.

- Skill: `search-atlas-group/case-study-engine` (Agent Skill)
- Install (CLI): `npx get-skill add search-atlas-group/case-study-engine`
- Raw SKILL.md: https://api.skillmd.com/api/skills/search-atlas-group/case-study-engine/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: search-atlas-group (https://skillmd.com/u/search-atlas-group)
- Updated: 2026-11-11
- Page: https://skillmd.com/skills/search-atlas-group/case-study-engine

---


# Case-study engine

## Purpose

Find repeatable client outcomes, assemble the evidence, and draft a case study that marketing or sales can use after approval. This is not a report generator; it is an evidence and permission workflow.

## Required evidence

- Baseline and comparison period.
- Observed change and exact source.
- Intervention or scope delivered.
- Attribution limits and plausible alternatives.
- Client identity and permission status.
- Reviewer and publication decision.

## Method

1. Search supplied redacted results for meaningful, repeatable outcomes.
2. Create a candidate record using `templates/candidate.example.md`.
3. Separate observed facts from interpretation.
4. Draft a proof story using `report-writer` or `html-reports` only after the evidence record is reviewable.
5. Route client naming, publication, outbound sharing, and sales use to a named human approver.

## Guardrails

Do not infer causation from before/after change alone. Do not expose a client, publish, or send collateral automatically. Missing evidence remains unknown.

**Evidence:** transcript JSON `i` segments 459–468.

