# Dataset Intelligence

> Route any dataset, custom indexing, traceability, provenance, lineage, synchronization, or retirement request through one governed workflow. Use when a user points AIWG at data and wants an outcome without knowing schema or indexing terminology.

- Skill: `jmagly/dataset-intelligence` (Agent Skill)
- Install (CLI): `npx skillmds@latest add jmagly/dataset-intelligence`
- Raw SKILL.md: https://api.skillmd.com/api/skills/jmagly/dataset-intelligence/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: jmagly (https://skillmd.com/u/jmagly)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/jmagly/dataset-intelligence

---


# Dataset Intelligence Router

Ask only for the source path or URI and desired outcome if they are not already
present. Do not require the user to choose a schema, adapter, backend, index, or
provenance vocabulary.

1. Invoke `dataset-intake` and return a concise recommendation containing:
   assumptions, rationale, privacy/locality/network implications, materialized
   artifacts, and safe defaults.
2. Route assessment to `dataset-source-assess`, capability composition to
   `dataset-capability-recommend`, and approval preparation to
   `dataset-plan-review`.
3. Route an approved plan to `dataset-ingest`; route inspection to
   `dataset-trace` or `dataset-verify`; route portable output to
   `dataset-export`; route removal to `dataset-retire`.
4. Carry stable intake, plan, run, checkpoint, approval, verification, and
   evidence references between phases. Never rely on conversation state as the
   record of authority.

Every operational step delegates to `aiwg dataset`. This router must not read a
source, execute a connector, write an index, advance a checkpoint, or delete
data. Indexes and Fortemi shards are derived, never canonical persistence.

SDLC, research, knowledge-base, media, marketing, ops, and project-local callers
all use the same intake and handoff contracts; domain additions are namespaced.