Artifacts: before and after
Forge should feel like better packaging of current work, not a second taxonomy. Use this table when translating existing artifacts and tooling.
| Familiar artifact | Forge-oriented equivalent | Notes |
|---|---|---|
| User story (refined, acceptance criteria) | Ingot (spec-shaped backlog item) | Keep one hierarchy; refine intent before large agent or team execution (spec-driven) |
| Task / sub-task / dev card | Spark | 1:1 with WBS task IDs—no duplicate task system |
| Sprint board / Kanban columns | Charge (view) + your tracker | Charge is today’s committed Sparks and decisions, not every possible column |
| Sprint Review output / release checklist | Assay Gate pack | Tie done to criteria, tests, and traceable IDs (ceremonies) |
| Decision log / steering notes | Ember Log, ADR, directive | Directional and cross-cutting calls stay named and linkable in the repo |
| Wiki architecture page | ADR + Architecture blueprint sections | Prefer versioned records next to code |
| Status email / slide deck | Journal / forge log (where used) | Log decisions, not narration—automate routine journals when possible |
Vocabulary detail: Forge naming reference. Process map: Forge process and flows.
Minimal starting set (from pilot practice): milestone/epic narrative, story + acceptance criteria, ADR when a choice is cross-cutting, tasks/PRs tied to work-unit IDs, tests tied to criteria, release notes referencing those IDs.
Next: Sponsors hub · Governance and evidence