Show HN: LLM-wiki – One command Karpathy's wiki with QMD search for Claude/Codex (github.com)

A new tool named llm-wiki has been announced, enabling developers to create and maintain project-specific wikis using AI models like Claude Code and Codex (GPT-5.5). This innovative agent skill streamlines knowledge management by offering three key workflows: bootstrap-wiki for establishing a project-local knowledge base, wiki-researcher for querying past project information, and wiki-plan for planning implementations based on historical context. By utilizing advanced semantic search capabilities through QMD, llm-wiki enhances the efficiency of project planning and execution. The significance of llm-wiki for the AI/ML community lies in its potential to optimize project management and knowledge sharing. By integrating with Claude Code and Codex, it allows users to leverage historical data without reinventing the wheel, ultimately improving productivity. The tool's design ensures that documentation is grounded in real code practices, as it pulls information from source files and transparently notes any gaps in knowledge. This approach not only fosters a culture of continuous learning but also aligns well with the agile development methodologies frequently employed in AI projects, making it a valuable asset for developers and teams focused on AI/ML initiatives.