Unfudged.io NEW

UNF filesystem flight recorder that continuously captures file-system snapshots and enables point-in-time restores.Records changes per second, supports macOS fsevents and Linux inotify, respects.gitignore and skips binaries.

Content-addressed storage uses blake3 hashing and SQLite ACID guarantees to deduplicate content and keep a local-only store.CLI and desktop app provide commands like unf watch, unf log, unf diff, unf restore and unf recap for timeline exploration and file-level rewind.

Automatic retention policy enforces 24-hour full, 7-day hourly, and 30-day daily decay while keeping storage outside project directories.Common use cases include recovering from AI agent mass-refactors, accidental deletions of env or migration files, restoring overwritten fixtures, and retrieving intermediate versions not captured by git.

Key Features

• Continuous filesystem flight recorder that captures snapshots for point-in-time restores
• Per-second change recording using macOS fsevents and Linux inotify; respects .gitignore and skips binaries
• Content-addressed storage with blake3 hashing and SQLite ACID-backed deduplication in a local-only store
• CLI and desktop app exposing commands (unf watch, unf log, unf diff, unf restore, unf recap) for timeline exploration and file-level rewind
• Automatic retention policy (24-hour full, 7-day hourly, 30-day daily) with storage kept outside project directories

Use Cases

• Instantly restore accidental edits or deletions by time-traveling files to per-second snapshots with unf's point-in-time restore, local-only recovery, and low-resource blake3+SQLite deduplication
• Maintain continuous, low-overhead versioning for development projects—revert code or configs to any per-second state, explore timelines via CLI/desktop, and minimize storage using content-addressed deduplication and automatic retention policies
• Perform file-level forensic investigation and auditing by exploring filesystem timelines captured from macOS fsevents or Linux inotify, recovering exact content-addressed versions, and enforcing retention for compliance

Who is it for?

• Software developers
• Machine learning engineers
• Open-source contributors
• Independent developers
• Coding assistants