2026-08-01 — Project kickoff

Founding session. From the initial idea ("a manager for architectures, projects and frameworks") to a named project with a study and recorded decisions.

What happened

1. Viability analysis: the gap is the middle ground between GitHub Templates (dumb copy) and Backstage (heavyweight enterprise portal) — a personal/lightweight template manager with repo creation.
2. Engine study (two passes: Android/KMP first, then generalized to multi-platform) → study/engine-v1.md, with requirements, state of the art, 9 decisions, edge cases, risks and roadmap.
3. Key requirement change: multi-forge (FR9) — the target repo is created on any git host, not only GitHub → ADR-0009.
4. Own engine vs Copier: current state of the art verified; we build our own → ADR-0001.
5. Name: ~70 candidates verified across 6 rounds (GitHub + npm + product-collision searches). Winner: Templetry → ADR-0008. Fully-available runners-up kept as plan B: Repojig, Forgehand.
6. Infrastructure: GitHub CLI installed; this wiki repo created with the study split into ADRs; templetry GitHub organization created; wiki published and translated to English.

State at close

• Accepted decisions: 0001–0005, 0008, 0009. Open: 0006 (language), 0007 (directive grammar).
• Immediate pending item: reserve templetry on npm (verified availability is not a reservation).

Next steps (Phase 0)

• Dry-run manifest validation: hand-write the template.yml of 3 dissimilar templates (KMP, React, FastAPI) against the draft in spec/template-yml.md.
• Close ADR-0007 (directive grammar) with what the validation teaches.
• Close ADR-0006 (language) before prototyping Phase 1.