# Dataset Intake > Turn a source path or URI and desired outcome into a governed dataset intake envelope for any AIWG domain. - Skill: `jmagly/dataset-intake` (Agent Skill) - Install (CLI): `npx skillmds@latest add jmagly/dataset-intake` - Raw SKILL.md: https://api.skillmd.com/api/skills/jmagly/dataset-intake/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-intake --- # Dataset Intake Accept a source path/URI and desired outcome. Default the calling domain to `project-local`; SDLC, research, knowledge-base, media, marketing, and ops use the same envelope and may add only namespaced extensions. Create a `dataset-intake/v1` envelope without opening the source. Record opaque source locator, outcome, domain, known privacy/locality/network constraints, authorization references, and independently requested capabilities (`search`, `traceability`, `provenance`, `graph`, `export`). Do not infer credential values. Then propose `aiwg dataset preview --json` and hand the intake reference to `dataset-source-assess`. For a question about existing normalized AI session history, route inspection to `session-explore` before proposing a new dataset. A separately requested export or derived index can enter this intake using its approved source/evidence references, privacy constraints and intended outcome. Do not scan provider homes or treat catalog inspection as approval to copy transcripts into another store.