Stars
Forks
Watchers
Siftly
Siftly transforms your Twitter/X bookmarks from a chaotic pile of saved tweets into a searchable, AI-categorized knowledge base with an interactive visual mindmap. With over 2,700 GitHub stars since March 2026, the platform runs a four-stage enrichment pipeline on each bookmark: entity extraction mines hashtags, URLs, @mentions, and 100+ known tool domains without API calls; vision analysis generates 30-40 visual tags per image using the Anthropic SDK; semantic tagging produces 25-35 searchable descriptors; and categorization assigns one to three categories with confidence scores. Search combines SQLite FTS5 full-text indexing with Claude-based semantic reranking, narrowing candidates through keyword matching, category-intent detection, and deduplication before sending a bounded set for LLM relevance scoring, letting you find bookmarks by meaning rather than exact keywords. The interactive mindmap built on @xyflow/react renders your entire collection as a force-directed graph organized by category with expandable nodes, color-coded legends, and direct links to original tweets. Import bookmarks through a built-in bookmarklet or console script without browser extensions, then browse in grid or list view with filters for category, media type, and date range. Export as CSV, JSON, or category-grouped ZIP archives. Prisma 7 manages the local SQLite database with FTS5 built in, requiring zero external database setup. A bundled CLI provides JSON-output commands for stats, search, and category management. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.
Benefits
- AI-Powered Semantic Bookmark Search
- Find bookmarks by meaning, not keywords — FTS5 candidate narrowing feeds a bounded set to Claude for semantic reranking, surfacing results like "funny meme about crypto crashing" from raw tweet content.
- Four-Stage Enrichment Pipeline
- Each bookmark passes through entity extraction, vision analysis generating 30-40 tags per image, semantic tagging producing 25-35 descriptors, and categorization with confidence scores — all running as a background SSE job.
- Interactive Mindmap Visualization
- Explore your entire bookmark collection as a force-directed graph built on @xyflow/react, with expandable category nodes, color-coded legends, and direct click-through links to original tweets on X.
- Privacy-First Local Architecture
- All data stays in a local SQLite database with zero external dependencies beyond configured AI API calls — no telemetry, no tracking, no third-party data sharing, and full export to CSV, JSON, or ZIP.
Features
- Vision Analysis
- Anthropic SDK reads text, objects, and context from every image, GIF, and video thumbnail, generating 30 to 40 visual tags per media item for searchable enrichment.
- Extensionless Bookmark Import
- Import Twitter/X bookmarks via a built-in bookmarklet or console script without installing any browser extension, with automatic duplicate detection on tweet IDs.
- SQLite FTS5 Search
- Full-text search powered by SQLite FTS5 with category-intent regex matching, date filtering, and media-type filtering, feeding bounded candidate sets to LLM reranking.
- Multi-Format Export
- Download your organized collection as CSV, JSON, or category-grouped ZIP archives using JSZip, with the bundled CLI providing JSON-output commands for programmatic access.
- Force-Directed Mindmap
- Interactive graph visualization using @xyflow/react displays bookmarks organized by category with expandable nodes, inline AI categorization triggers, and color-coded legends.