fix: add missing threading import in statistical_detector

chore: VERSION 7.7.0 -> 7.8.0, update version assertions

fix: sanitize numpy types, forensics isolation, admin warning, docstring, remove dotenv

fix: DIRE confidence fragile check and missing available flag

perf: replace O(n^2) radial FFT loops with vectorized np.bincount

fix: KL divergence missing fftshift and ELA neutral score for lossless

fix: three thread-safety bugs — global random seed, race condition, unseeded sampling

fix: inter-block blockiness and missing method keys in advanced_ai_detector

fix: double-stacking boost and wrong blockiness metric in ai_detector

fix: resolve model/database paths relative to __file__ not CWD

fix: CLIP detection always disabled — db_available never set True

feat: Noiseprint camera fingerprint and CFA Bayer artifact detectors

fix(phase-20): update detection_version and methods_used assertions

fix(phase-20): update 15 hardcoded signal-count assertions 26 -> 28

fix(phase-20): wire jpeg_ghost and noise_map into ensemble scoring

fix(phase-20): add missing jpeg_ghost_detector.py and noise_map_detector.py

feat: JPEG Ghost + Noise Map detectors — signals 27 & 28

Merge branch 'feature/webhook' of https://github.com/abinaze/VeriFile-X into feature/webhook

feat: fix webhook suspend bug and CodeQL clear-text logging alert

Potential fix for pull request finding 'CodeQL / Clear-text logging of sensitive information'

feat(feature): webhook delivery system with HMAC signing and retry

test: update version assertions from 7.1.0 to 7.2.0

security(admin): validate X-Admin-Key against SHA-256 hash, not just length

fix(ensemble): resolve XGB path via __file__; log when own_weight excluded

fix(batch): fix uncertain_count negative values with exclusive classification

fix(prnu): replace uniform_filter fallback with Laplacian noise residual

fix(ela): halve confidence for lossless PNG/WebP inputs

fix(models): remove FileInfo.size field with no matching key in report dict

fix(config): replace os.getenv() defaults with proper pydantic-settings fields

fix(analyze): remove duplicate Allowed: string in 415 error

fix(paths): resolve all storage files relative to package, not CWD

fix(paths): resolve all storage files relative to package, not CWD

fix(api-keys): add write lock to verify_key; resolve path via __file__

fix(dct): assign smooth_score in else-branch — prevents NameError on small images

fix(cache): remove double TTL check; wrap set() in RLock

fix(ci): fix 10 test failures from previous commits

fix(stability): pywt level clamp + CLIP None centroid guard

security: add HSTS/CSP/Permissions-Policy headers + auth on metrics reset

fix(concurrency): audit log write lock + timestamped rotation backup

fix(accuracy): fix natural model size bug + deterministic KL divergence

fix(concurrency): move cache TTL check inside threading.Lock

fix(accuracy): replace deprecated PIL._getexif() in metadata_forensics

security(codeql): fix information exposure #8 #19 #20 #21

fix(deps): python-multipart 0.0.27->0.0.26 (0.0.27 does not exist on PyPI)

fix(async): correct asyncio API + latency variable reference

security(deps): python-multipart 0.0.22->0.0.27, starlette lock bump

security+fix: torch.load weights_only, cache thread safety, sse cleanup

fix(accuracy): fix ELA Signal 3 tautological percentile

fix(perf): cache natural spectrum model in StatisticalDetector

fix(audit): use live VERSION from settings, add 50MB rotation cap