The GPU detection in gpu.py had two issues preventing GPU usage in
containers (especially rootless podman with CDI):
1. When torch was installed but torch.cuda.is_available() returned
False, the function returned immediately without trying the
ONNX Runtime + nvidia-smi fallback. This meant a CPU-only torch
build (installed before GPU was available) would block all GPU
detection, even for ONNX-based tools.
2. The failure logged a generic "torch loaded but CUDA not available"
with no diagnostic information, making it impossible to debug
whether the issue was a CPU-only build, missing libraries, or
device permissions.
The fix restructures gpu_available() into three detection tiers
(torch -> ONNX Runtime -> nvidia-smi) that always fall through on
failure. When torch CUDA fails, it now checks torch.version.cuda to
distinguish CPU-only builds from CUDA builds that can't access the
GPU, and logs LD_LIBRARY_PATH, torch.cuda.init() errors, and
nvidia-smi results.
Also fixes two env var passthrough bugs in buildMinimalEnv():
- SNAPOTTER_GPU was never passed to the Python subprocess, so the
user-facing GPU override env var had no effect
- MODELS_DIR was a dead entry (never set as env var); replaced with
MODELS_PATH which the Dockerfile sets and Python scripts read
Closes#134
Auth: login rate limit 30/min (was 500), global rate limit 1000/min (was
unlimited), password/username max lengths on all Zod schemas, session
invalidation on role change, API key legacy scan bounded to 100 keys.
SVG: hardened regex sanitizer with CDATA stripping, XML entity decoding,
set/animate/iframe/embed blocking, comprehensive data: URI blocking,
use element external href blocking. 11 attack payload fixtures added.
SSRF: fixed DNS rebinding TOCTOU by pinning resolved IPs via custom
HTTP/HTTPS agents. Added 6to4 and NAT64 to blocked IPv6 ranges.
Docker: capability dropping (cap_drop ALL + minimal cap_add), resource
limits (4g/8g mem, 512/1024 pids), healthcheck timeout, password
removed from startup banner, default password warning comments.
Network: CSP and HSTS applied in all environments (not just production),
stack traces removed from all error responses, internal paths stripped
from error details, per-route rate limits on uploads (60/min) and URL
fetches (200/hour).
Files: exclusive temp file creation (O_EXCL), disk space circuit
breaker, per-user storage quotas, settings payload 64KB size guard.
Python sidecar: script name allowlist in dispatcher, minimal environment
for subprocess spawns.
Dependencies: fixed 6 production CVEs (drizzle-orm, fastify, fast-uri,
@fastify/static, next, archiver/lodash). Pinned all GitHub Actions to
SHA hashes.
114 security tests added. Full OWASP Top 10 penetration test matrix
verified against production Docker container (30/30 pass after
hardening).
- Replace content-aware-crop with ai-canvas-expand in TOOLS[], AI_TOOL_IDS,
and FEATURE_BUNDLES (matching the already-updated tool-registry.tsx and
feature-manifest.json from commit c6a5d3f)
- Fix trailing syntax error in features.ts (extra closing brace)
- Add ai-canvas-expand-settings mock to tool-registry test files
- Update watermark-image tests to expect 400 (validation rejection) instead
of 422 (processing failure) for corrupted image buffers, matching the
actual route behavior where validateImageBuffer catches them first
AVIF (and other Sharp-native formats) were written as raw bytes to a
.png temp file, causing PIL to fail with "cannot identify image file".
Every other AI module wrapper already converts via sharp().png().toBuffer()
before writing; face-landmarks was the only one that skipped this step.
Background images, device frames, custom shadows, and watermark text
were not rendering in the right-pane preview. The preview now updates
in real time for all settings: gradient/solid/image backgrounds, macOS/
Windows/Browser frame chrome, iPhone/MacBook/iPad frame indicators,
custom shadow parameters, and watermark text overlay.
Also fixes a React StrictMode effect-ordering race where the parent
tool-page reset cleared preview state set by the child Settings
component on initial mount.
- Refactor use-tool-processor and use-pipeline-processor hooks
- Enhance dropzone component with improved UX
- Improve seam carving with better error handling and tests
- Add JXL format encoding support to format-encoders
- Update tool routes for consistent format handling
- Add dropzone unit tests
The binary search found the right quality but sharp(buffer).toBuffer()
re-encoded at default quality 80, inflating the output (e.g. 50KB target
producing 90KB). Replaced buffer-wrapping with proper Sharp pipelines
that include .toFormat() with the proven quality. Also added progressive
dimension reduction when quality alone cannot reach the target, and
tightened tolerance to only accept at-or-below-target results.
The subdirectory removal test used setImmediate ticks to wait for async
cleanup, which was insufficient on slow CI runners. Replaced with real
setTimeout delays (100 x 10ms) to give async readdir/stat/rm I/O time
to complete.
Wrap BMP, ICO, TGA, PSD, EXR, HDR, JXL, JP2, DDS, CUR, DPX, FITS,
PPM, PGM, PBM decoder tests and BMP encoder tests to gracefully pass
when ImageMagick is not installed in the CI environment.
Tests verify well-exposed and bright-but-normal images are not
darkened, all modes produce valid output, and intensity slider
produces visibly different results.
CLAHE provides adaptive local contrast, normalise stretches the
histogram, and gamma adjusts exposure perceptually. Replaces the old
modulate/linear pipeline that compounded errors and darkened images.
Preset multipliers now include clahe and normalise entries.
Key fixes beyond the spec:
- maxSlope rounded to integer (Sharp requirement)
- White balance uses linear() instead of recomb() to avoid float-cast
that breaks CLAHE in the libvips pipeline
- CLAHE tile size adapts to image dimensions (1x1 for tiny images)
- Gamma clamped to Sharp's valid range (1.0-3.0)
- Normalise lower/upper correctly mapped to percentile cutoffs
Contrast score now uses linear stdevLum/1.2 centered at 50 (was
miscalibrated 25-75 range centered at 75). Corrections use dead zones
(score 40-60 = zero) so well-exposed images get near-zero adjustments
instead of being darkened.
- fix WebP export silently producing PNG when background is non-transparent
- fix autosave not converting blob: URLs inside image-type canvas objects
- fix project load not resetting selection/crop/clipboard state
- fix rotateCanvas not updating object rotation attributes
- fix flipCanvas not negating object rotation attributes
- fix line shadow props overridden by effect spread ordering
- fix "outside" stroke position rendering same as "center"
- add missing pencil tool keyboard shortcut (N)
- remove misleading resample dropdown from image resize dialog
- fix E2E autosave tests for production builds (no Vite dynamic imports)
- fix color picker test case sensitivity (CSS uppercase vs DOM text)
- add 4 unit tests for rotation attribute transforms
Canvas rendering:
- Fix Konva filter application order (filters before cache)
- Implement 6 missing filters (motionBlur, radialBlur, surfaceBlur, vignette, grain, sharpen)
- Implement exposure, vibrance, warmth adjustments as custom Konva filters
- Apply layer blend modes via globalCompositeOperation
- Apply object effects (drop shadow, outer glow, stroke) to all shapes
- Mount SmartGuidesOverlay during move tool drag
- Clip pixel grid to visible viewport (200-line cap for performance)
Store logic:
- resizeImage now scales all objects proportionally (points, radii, fontSize)
- rotate/flip/trim handle line/arrow points arrays and center-based objects
- applyCrop creates cropped source image via offscreen canvas
- invertSelection creates mask from bounds when no mask exists
- cutObjects uses single atomic set() to prevent race conditions
- sendToBack respects layer ordering in multi-layer documents
- Add batchNudge() and commitHistory() for undoable nudge operations
- Add updateLayerThumbnail() method
Tool hooks:
- Fix clone stamp/dodge/burn perf (toDataURL only on mouseUp, not every move)
- Fix magic wand zoom/pixelRatio with explicit stage.toCanvas() viewport
- Fix eyedropper sampling with unzoomed canvas export
- Fix selection tool stale closure via isDrawingRef
- Implement polygonal lasso (click-to-place vertices, double-click to close)
- Implement selection subtract mode (geometric and mask-based)
- Implement gradient live preview during drag
- Fix transform/move tool to persist changes and handle ellipse/polygon/star
UI wiring:
- Mount rulers and guidelines in editor page
- Wire histogram with live canvas imageData
- Wire autosave recovery with blob-to-dataURL conversion
- Wire fill dialog to Shift+Backspace shortcut
- Wire eyedropper and transform options to options bar
- Fix history panel undo/redo button reactive state via useSyncExternalStore
- Fix layer row name click to select layer (timer-based click/dblclick)
- Fix zoom animation coordinate drift with progressive store sync
- Fix copy merged to use Konva stage composite export
Tests:
- 49 new unit tests (store fixes + konva filters)
- 8 new E2E test files with 39 test cases
Add colorBlindness() operation with 8 simulation matrices (Vienot/Machado)
for protanopia, deuteranopia, tritanopia, protanomaly, deuteranomaly,
tritanomaly, achromatopsia, and blue cone monochromacy.
fix: add mediapipe to upscale-enhance bundle for face enhancement
The enhance-faces tool requires MediaPipe for face detection, but the
upscale-enhance feature bundle did not include mediapipe in its pip
packages. Added mediapipe to both amd64 and arm64 package lists.
Added 25 regression tests validating bundle dependency completeness.
Closes#129
The enhance-faces tool requires MediaPipe for face detection, but the
upscale-enhance feature bundle did not include mediapipe in its pip
packages. Users who installed only the upscale-enhance bundle got
"Face detection requires MediaPipe" errors. Added mediapipe to both
amd64 and arm64 package lists, matching the pattern used by the
face-detection and photo-restoration bundles.
Also added feature-manifest.test.ts with 25 tests validating bundle
dependency completeness to prevent similar missing-dependency bugs.
Closes#129
The default-view redirect in HomePage fired on every mount, not just the
initial page load. A module-level flag now gates the redirect so it only
applies once per session, allowing users to switch to sidebar view when
grid is the default.
Closes#128
SnapOtter