feat: M1 foundation packages + trip-cluster app #1

2026-06-27T19:07:31+02:00

m038 commented

2026-06-27 19:07:31 +02:00

What

Implements the full M1 plan (docs/superpowers/plans/2026-06-27-immich-photo-flow-m1.md): the photoflow shared monorepo and the trip-cluster Flask app end-to-end — ingest -> density-adaptive clustering -> review UI -> idempotent write-back to Immich.

shared/photoflow/immich — the one Immich REST client + _pipeline/ tag conventions
shared/photoflow/core — SQLite store + domain models (the only SQL owner)
shared/photoflow/ui — base template, DaisyUI/Tailwind/Alpine/HTMX, shared grid+lightbox
apps/trip-cluster — argparse CLI (ingest|cluster|serve|apply) + master/detail review UI on port 8084

All 17 plan tasks shipped, built TDD (failing test first per unit). Immich is the source of truth; SQLite is a rebuildable working layer; write-back is idempotent (writeback_log) and confirmation-gated.

Tests

66 green: 60 unit/route (pytest + pytest-httpserver, Immich mocked) + 6 Playwright UI. Dev env runs Python 3.14; the Docker/target runtime stays 3.12; all pinned deps install on both.

Review

A 6-reviewer pass (correctness, security, reliability, maintainability, testing, project-standards) ran on the branch. No exploitable security issues. Nine findings fixed with regression tests:

P1 — upsert_tag failure no longer aborts the whole apply_all batch (moved inside the per-asset try/except)
P2 — processed-tag write failures surfaced in the result; confirmation gate on single-cluster apply; approve-high-confidence count fixed; DB connection closed even if a route raises; incremental-ingest watermark no longer strands thumbnail-failed assets
P3 — split_cluster empty-cluster guard; writeback_log index; ingest .tmp cleanup

Remaining items are report-only cleanups and do not block merge.

🤖 Generated with Claude Code

## What Implements the full **M1** plan (`docs/superpowers/plans/2026-06-27-immich-photo-flow-m1.md`): the `photoflow` shared monorepo and the `trip-cluster` Flask app end-to-end — ingest -> density-adaptive clustering -> review UI -> idempotent write-back to Immich. - `shared/photoflow/immich` — the one Immich REST client + `_pipeline/` tag conventions - `shared/photoflow/core` — SQLite store + domain models (the only SQL owner) - `shared/photoflow/ui` — base template, DaisyUI/Tailwind/Alpine/HTMX, shared grid+lightbox - `apps/trip-cluster` — argparse CLI (`ingest|cluster|serve|apply`) + master/detail review UI on port 8084 All 17 plan tasks shipped, built TDD (failing test first per unit). Immich is the source of truth; SQLite is a rebuildable working layer; write-back is idempotent (`writeback_log`) and confirmation-gated. ## Tests 66 green: 60 unit/route (pytest + pytest-httpserver, Immich mocked) + 6 Playwright UI. Dev env runs Python 3.14; the Docker/target runtime stays 3.12; all pinned deps install on both. ## Review A 6-reviewer pass (correctness, security, reliability, maintainability, testing, project-standards) ran on the branch. No exploitable security issues. Nine findings fixed with regression tests: - **P1** — `upsert_tag` failure no longer aborts the whole `apply_all` batch (moved inside the per-asset try/except) - **P2** — processed-tag write failures surfaced in the result; confirmation gate on single-cluster apply; `approve-high-confidence` count fixed; DB connection closed even if a route raises; incremental-ingest watermark no longer strands thumbnail-failed assets - **P3** — `split_cluster` empty-cluster guard; `writeback_log` index; ingest `.tmp` cleanup Remaining items are report-only cleanups and do not block merge. 🤖 Generated with [Claude Code](https://claude.com/claude-code)

m038 added 19 commits 2026-06-27 19:07:31 +02:00

feat: monorepo scaffold — photoflow shared dist + trip-cluster app skeleton 01ac95059e

feat(immich): ImmichClient read — search_assets, list_tags, resolve_tag_id, download_thumbnail 862480916a

feat(core): domain models + Store (schema, connection, meta) 4db684fbdc

feat(ui): base.html, macros (badges, lightbox), shared.js, register_shared_ui c36d93e3c7

feat(immich): write-back (upsert_tag, tag_assets) + _pipeline tag conventions da5b1de2a2

feat(core): asset & tag data-access 351a001903

feat(trip-cluster): density-adaptive timestamp clustering + tag seeds + anchors 76be1fb5af

feat(core): cluster data-access — attention sort, neighbors, split/merge, writeback log 9ec6512428

feat(trip-cluster): config, factory, CLI skeleton, health + thumb proxy e01b30a5b5

feat(trip-cluster): coverage detection — completeness flags + outliers 8d58ec03a9

feat(trip-cluster): ingest — scopeable, incremental, processed read-back 6d0c9662ee

feat(trip-cluster): cluster review operations (approve/non-trip/skip/split/merge/bulk) a9b443cd53

feat(trip-cluster): cluster orchestration (seed derivation + coverage) + CLI 5256df4117

feat(trip-cluster): idempotent write-back + CLI apply with confirmation 44c05444d3

feat(trip-cluster): master/detail review routes + templates + clusterReview JS edf0735098

test(trip-cluster): Playwright UI — grid nav, lightbox, cluster switching, approve c4ff5bcb06

docs+docker: README, CLAUDE.md, Dockerfile/compose (8084), M1 validation gate d67b677b6a

fix(review): harden write-back, confirmation gate, and edge cases ae093881fa

- write-back: call upsert_tag inside _apply_tag's try so a tag-create failure
  is caught per-asset and apply_all no longer aborts mid-batch (was P1)
- write-back: surface _pipeline/processed write failures in the result instead
  of discarding them (was reported as success)
- ui: add title to the high confidence badge so approve-high-confidence's
  pre-action count is non-zero; confirm() before single-cluster apply
- core: Store context manager; close DB connection even if a route raises;
  guard split_cluster against a first-member boundary (empty cluster); add
  writeback_log lookup index
- ingest: split thumbnail download/write error handling and clean up the
  .tmp file on a write failure
- tests: upsert/processed write-failure regression tests + split-guard test

fix(ingest): don't advance incremental cursor past a failed/sampled asset bd8fe59b49

A later successful asset could push last_ingest_at past an earlier asset whose
thumbnail download failed, permanently excluding it from later incremental runs
(only --full recovered it). Now the cursor never advances to/past the earliest
failed asset, never below the prior cursor, and not at all on a --subset run.
Regression tests added. (review finding #9)

m038 merged commit d5b073da9c into master

2026-06-27 19:10:50 +02:00

Sign in to join this conversation.

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: m038/immich-photo-flow#1