What problem does it solve?
UX practitioners often lose evidence provenance when turning research into durable documents, producing artifacts that blur observed facts, reported claims, and assumptions. This Skill produces one current, evidence-labelled Markdown asset per mode so engineers, designers, and agents can consume UX decisions without inventing research or losing uncertainty.
Core Features & Use Cases
- Five bounded modes: frame-needs, map-journey, sketch-structure, decide-inclusion, and prepare-handoff, each producing a single canonical Markdown asset under
.copilot-tracking/ux-artifacts/.
- Evidence contract: every output separates Observed, Reported, Assumed, and Unresolved sections and ends with a mandatory human-review gate that only a person can check.
- Destination mapping: optional Figma, FigJam, or Mural intent generation that shapes content for those tools without authenticating or writing externally.
- Use Case: A designer with completed user interviews invokes map-journey mode with the interview notes as source, receiving a staged journey asset where every pain point carries its evidence class and unresolved gaps stay explicit.
Quick Start
Ask the agent to use the ux-artifacts skill with mode map-journey, a project slug, a subject, and your research notes as the source to produce an evidence-labelled journey map.