# Meta-Process Capture — Expert Panel Dataset Builder

> Document the multi-agent expert panel methodology as a reusable system for building datasets that inform skills and systems. The process — assign framework experts to repositories, ask structured questions, cross-examine, synthesize, map to requirements, validate — produced high-quality sourced findings. Capture this as a repeatable pattern. **Key elements to document**: - Expert assignment protocol (one agent per source repo, source-code-only evidence standard) - Question group design (themed questions → cross-examination → synthesis) - Cross-examination as adversarial validation (experts challenge each other's claims) - Phased output (discussion → requirement mapping → synthesis → validation) - Traceability chain (claim → expert citation → file:line evidence) - Session continuity handling (state file enables cross-session resumption) **Input artifacts**: `plugins/plugin-creator/skills/assessor/references/ARL/ARL-agent-instructions.md`, `qa-expert-panel.md`

- Skill: `tools-only/meta-process-capture-expert-panel-dataset-builder` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds@latest add tools-only/meta-process-capture-expert-panel-dataset-builder`
- Raw SKILL.md: https://api.skillmd.com/api/skills/tools-only/meta-process-capture-expert-panel-dataset-builder/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: tools-only (https://skillmd.com/u/tools-only)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/tools-only/meta-process-capture-expert-panel-dataset-builder

---


**Suggested location**: New skill or methodology document — captures the meta-process, not the ARL content