Summary This project addresses X (formerly Twitter)'s spam and bot problem with a passive browser extension. It aims to enhance user experience by filtering unwanted content and offering insights, functioning as a public-good, open-source initiative.

How It Works The extension uses AI to passively scan X comments, identifying spam and adult bots. It implements "true blocking" via X's native API, not just hiding content. The system synchronizes community-maintained public block/whitelist lists from the project's GitHub repository. Architecture includes a Chrome MV3 extension (React, Tailwind) and a Cloudflare Worker backend (Hono, D1 SQLite) for list management and an admin interface.

Quick Start & Requirements

• Installation: Chrome Web Store. For others, download .zip release and load as unpacked extension.
• Prerequisites: Chromium browser. Development requires pnpm.
• LLM Configuration: OpenAI-compatible /chat/completions endpoint needed, configured via secrets (LLM_API_BASE, LLM_API_MODEL, LLM_API_KEY). ADMIN_TOKEN required for admin interface.
• Links: Chrome Web Store, GitHub Releases.

Highlighted Details

• AI Spam Shield: Automatically identifies and enables one-click blocking of spam/adult bots.
• Community Lists: Public block/whitelist lists in data/ sync every six hours; Git history provides audit trail.
• Privacy-First: No user data collection or registration.
• Planned Features: Future development includes KOL signal scoring, profile digests, social graph hints, and data export.

Maintenance & Community Contributions via PRs/issues guided by CONTRIBUTING.md and GOVERNANCE.md. Governance emphasizes transparency: logged actions, appeals via GitHub issues. Roadmap includes planned "Pillars" beyond current functionality.

Licensing & Compatibility Licensed under AGPL-3.0. This copyleft license requires derivative works to be shared under the same terms, potentially impacting commercial use or integration with closed-source systems.

Limitations & Caveats Only spam-filtering (Pillar 01) is live; other features are under development. AI moderation is strictly for commercial spam/adult bots, excluding opinions/politics. Requires external LLM API configuration. AGPL-3.0 license needs careful consideration for commercial adoption.