Show HN: MCP server for persistent context and safety in Claude Code (github.com)

AXME Code has launched a new MCP server designed to enhance Claude Code by providing persistent context and improved safety measures for software development sessions. This tool addresses common frustrations developers face, such as losing context after multiple sessions or inadvertently executing dangerous commands like `git push --force`. By maintaining a comprehensive knowledge base that includes project architecture, decisions, and safety protocols, AXME Code ensures that developers can seamlessly continue their work without redundant explanations or risks of unsafe operations. Significantly, AXME Code combines both context management and hard safety enforcement through pre-execution hooks that block harmful commands. This structured approach allows the AI agent to learn from previous sessions, automatically accumulate decisions and memories, and ensure compliance with safety rules. Developers can now utilize Claude Code more effectively, backed by an automatic auditing system that captures insights even if a session closes unexpectedly. The integration of AXME Code not only streamlines the coding process but also enhances the reliability and security of AI-assisted development workflows, making it a crucial advancement for the AI/ML community focused on building trustworthy software solutions.