0xwilliamortiz
Featured
openclaude-improved
OpenClaude is an open-source CLI coding agent that runs on any platform and supports a wide range of LLM providers, offering the same tools and workflows as Claude Code.
AI repository intelligence
FR-AI / ANALYSISWhy it is worth attention
It produces machine-readable intelligence (scored claims, knowledge graphs) rather than unstructured text, uses a stealth drone arsenal to bypass anti-bot defenses and paywalls, and integrates natively with AI assistants via MCP, all under an MIT license with self-hosting.
Who it is for
- AI developers building agentic research tools
- Data scientists and analysts needing verifiable, structured data
- Journalists and researchers who require multifaceted source gathering
- Developers integrating research capabilities into existing workflows
Use cases
- Enhancing AI assistants (Claude, ChatGPT) with real-time, cited research
- Competitive and market intelligence gathering from 70+ specialized APIs
- Academic literature review with automatic claim extraction and quality scoring
- Building persistent knowledge graphs that accumulate across research sessions
Strengths
- Dual-model extraction (fast Haiku for bulk, powerful Sonnet for enrichment) with per-claim quality scoring (0.55 threshold)
- Stealth drone fetch stack (8 layers) that bypasses Cloudflare, DataDome, paywalls, and rate limits
- 70+ API sources across 37 categories (academic, financial, government, security, etc.) for broad coverage
- Native MCP support plus REST/SSE streaming for real-time progress and integration with any MCP client
Considerations
- Requires your own LLM API key (costs $0.30–$2.00 per research mission in tokens)
- Full feature set relies on PostgreSQL, Neo4j, and optional Qdrant (self-hosted infrastructure)
- Stealth scraping may violate website terms of service; ethical use is the user's responsibility
README quick start
Quick Start
Description
Open-source deep research engine that builds structured knowledge graphs. MCP-native. avg 0.82+ quality.
Related repositories
Similar projects matched by category, topics, and programming language.
slvDev
Featured
esp32-ai
A 28.9 million parameter language model runs on an $8 ESP32-S3 microcontroller entirely on-device, generating simple stories at about 9.5 tokens per second.
gavamedia
Featured
deltafin
Deltafin is a research project that runs the 2.8-trillion-parameter Mixture-of-Experts model Kimi K3 on a single Apple Silicon Mac (e.g., M1 Max with 64 GB) at about 16 seconds per token, using exact, reproducible inference with local or streaming expert loading.