[fpliger]

Coordinator plan micro-PR. Records the user's D-4 decision (2026-10-19) in migration-tracking/plans/kilo-v2-migration-plan.md.

• §A10 D-4: answered. U-E1-01 part 1 is the kilo2 app directories (config, data, cache, state), set by the Kilo host. Part 2 (rebranding) is binary, user agent, theme, logo and user-visible strings, after D-1 and Q-DIST-01. kilo2 holds until storage consolidation (U-E8-02, after G3, D6).
• Options considered, with the reason each was rejected: upstream's opencode directories, v1's kilo directories, a subdirectory of v1's tree, renamed files in v1's directories, an environment-variable profile only. kilo2 is the only option that is both cheap and safe.
• Upstream contrast: upstream v2 upgrades v1 in place (same opencode directories, same opencode.db on release channels, V1Migration on first boot, opencode-next.db preview import, v1 config compatibility parsers). Kilo cannot do that while v1 IDEs run on v1 data (D5, D6). Reusable pieces named: v1 config parsers for U-E1-07/U-E3-03, the resumable one-time migration pattern for U-E8-02. The archived port's kilo2 preview identity is noted.
• U-E1-01 row: "config filenames" removed (done in feat(core): discover kilo.json and .kilo config directories #15018 under U-E1-07). Acceptance split into part 1 and part 2. Status notes that the app name and the 17 KILO_* env names landed with the host (feat(kilo-cli): add the Kilo host and the feature placement finder #15232).
• U-E1-07 row: owns config filenames. New acceptance item: project files (kilo.json, .kilo/) are shared with v1 whatever the app directory, so v2 never writes a key into them that v1 rejects.

No status, gate or other decision changes. Checks: bun script/upstream/progress.ts --check up to date. No Changeset.

Second commit: the upstream contrast is now explicit. Upstream's strategy: in-place upgrade, no coexistence. Why Kilo cannot use it until the end: the v1 IDEs run on the v1 store until G3 (D5), and D6 forbids converting or writing it while any v1 surface ships. When it becomes usable: at storage consolidation (U-E8-02, after G3), which should reuse upstream's one-time, resumable migration pattern.

[github-actions]

➖ v2 migration progress: 0.9% → 0.9% (0 done, 0 total, 0 dropped; 86/9829 cells)

Placement: 0 new marked lines in shared files, 0 new kilo_port annotations, 0 new core-internal imports.

Ratchets: 9 core-internal imports (base 9); shared patches by rung 29 seam, 1 patch (base 29 seam, 1 patch).

No change to the progress map.

Full map · how cells close

[kilo-code-bot]

Code Review Summary

Status: No Issues Found | Recommendation: Merge

Files Reviewed (1 file)

• migration-tracking/plans/kilo-v2-migration-plan.md - reviewed the new upstream-strategy and cutover-rationale paragraphs against packages/cli/src/database-path.ts, packages/core/src/database/v1-migration.bun.ts, packages/util/src/global.ts/global-roots.ts, packages/core/src/v1/config/, and the plan's own D5/D6, gate map, and U-E3/U-E8 rows; claims match the code and the plan is internally consistent. No images added, so no markdown image-format concerns.

Previous Review Summary (commit 2703aba)

Current summary above is authoritative. Previous snapshots are kept for context only.

Previous review (commit 2703aba)

Status: No Issues Found | Recommendation: Merge

Files Reviewed (1 file)

• migration-tracking/plans/kilo-v2-migration-plan.md - verified the D-4 answer, option table, and updated U-E1-01/U-E1-07 rows for internal consistency and factual accuracy against the referenced code (packages/core/src/database/v1-migration.bun.ts, packages/cli/src/database-path.ts, packages/core/src/v1/config/, #15018, #15232); no images added, so no markdown image-format concerns.

---

_{Reviewed by deepseek-v4.1-flash-1017 · Input: 49K · Output: 6.8K · Cached: 554.9K}

_{Review guidance: REVIEW.md from base branch v2}