AI / ML

DeusData released codebase-memory-mcp, a high-performance MCP server that indexes codebases into a persistent knowledge graph for AI coding agents. It uses tree-sitter AST analysis across 162 languages and hybrid LSP semantic resolution to answer structural queries in under a millisecond. The tool ships as a single static binary with zero dependencies, achieving full indexing of average repos in milliseconds and the Linux kernel in three minutes.

• Indexes codebases into a persistent knowledge graph for sub-millisecond structural query responses.
• Supports 162 languages using tree-sitter AST and hybrid LSP semantic type resolution.
• Delivers 99% token reduction compared to traditional context-passing methods.
• Distributes as a single static binary with zero runtime dependencies across macOS, Linux, and Windows.

Anthropic reports that its LLM, Claude, has discovered a previously unknown enzyme system featuring CRISPR-like repeats. This finding highlights the model's capacity for biological discovery and pattern recognition beyond standard text generation tasks. The announcement underscores the growing role of AI in accelerating scientific research and hypothesis generation.

• AI models are now capable of identifying novel biological mechanisms, not just analyzing existing data.
• The discovery involves a new enzyme system with CRISPR-like structural repeats.
• Anthropic is positioning Claude as a tool for active scientific discovery and hypothesis generation.
• This marks a shift from passive AI assistance to active contribution in life sciences research.