I gave Claude Code Unlimited Persistent Memory, and you can even talk to it!
Every AI coding agent — Claude Code, Gemini CLI, Cursor — forgets everything between sessions. They start blind, re-read files, re-discover bugs, and waste tokens rebuilding context you already gave them. I built a free, open-source MCP server that gives any CLI coding agent persistent, cross-session memory using Google's Gemini File Search API as a cloud-hosted RAG store. Here's how it works: at the end of each session, a structured summary is saved to a per-project Gemini store — what was built, what decisions were made, what bugs were found, and what's next. At the start of the next session, the agent retrieves semantically-searched context tailored to the task at hand. No flat files, no keyword matching — real AI-powered retrieval with configurable token budgets from 500 to 16,000 tokens. The result: your agent never starts from scratch. It picks up exactly where you left off, across machines, across sessions, for essentially zero cost. The tool works as both an MCP server (for Claude Code) and a standalone CLI tool (gmem) that works with any agent — Gemini CLI, Cursor, Aider, or anything else. 🔗 Links: Gemini Memory MCP Server (GitHub): https://github.com/alexcelewicz/CLI-Gemini-Unlimited-Memory Apportrait links to MCP and Sherlock Skill: https://apportrait.com/skills/4 https://apportrait.com/skills/3 Google Gemini API Key: https://aistudio.google.com/apikey Claude Code: https://docs.anthropic.com/en/docs/claude-code 💡 What this tool does: - Saves structured session summaries to per-project Gemini File Search stores - Retrieves relevant context via semantic search at session start - Two-phase retrieval: broad project context + task-specific queries - Works as MCP server (Claude Code) or standalone CLI (any agent) - Supports git hook auto-save for automatic session capture - Cloud-hosted — works across machines, survives compaction - Storage is free, indexing costs ~$0.15 per million tokens #ClaudeCode #AIMemory #CodingAgent #GeminiAI #MCPServer #DevTools #OpenSource #AIAutomation #PersistentMemory #RAG