Show HN: CogniKernel; AI Coding with cross session and cross platform memory (github.com)

CogniKernel has introduced a new tool for AI coding that enhances memory across coding sessions and platforms, designed for Claude Code and Codex. This tool captures significant decisions, constraints, and abandoned strategies during a coding session, storing them in a structured, event-sourced database that allows for effortless recall. Rather than relying on large language models (LLMs) to process memory, CogniKernel uses two lightweight fine-tuned encoder models for real-time classification and consolidation of relevant information, running locally to ensure data privacy and efficiency. This development is significant for the AI/ML community as it provides a solution that optimizes coding workflows without incurring cloud costs or latency associated with traditional memory tools. By minimizing unnecessary redecisions and read operations, CogniKernel enhances productivity, particularly in projects with complex evolving states. With features like cross-tool memory, automatic capture and recall, and fail-open reliability, it streamlines sessions significantly, allowing developers to focus on their work rather than on rediscovering prior decisions. The tool's design also emphasizes local operation and clear degradation paths, promising a dependable experience even in the case of errors.