Project memory for long-running AI work across agents, models, and sessions. Keep context, decisions, progress, and next steps in local project files.

AI agent skill that remembers every technical decision & bug fix across sessions — and learns from them. v4.0, MIT. | 跨会话记忆的AI编程助手知识大脑

Redis integration for Google Agent Development Kit (ADK) - Memory, Sessions, Search Tools, MCP

Persistent memory for Cursor IDE — like claude-mem, but for Cursor. Zero-config, no API key.

Shared memory and identity for your AI agents — Git-native, self-hosted, no vendor lock-in

redact + memory — industrial-grade Claude Code infrastructure. Secret protection + persistent session memory in one plugin.

Capture and restore high-value session memory for coding projects.

Cross-session conversation trail for Claude Code — lightweight hooks, zero deps, auto-injected on session start. 给 Claude Code 的跨会话留痕：纯 hook、零依赖、新会话自动回注。

Portable session-memory system for agents: provenance-carrying session evidence and queryable read models, with Codex as the current production adapter.

Claude skill that carries session state — goals, decisions, blockers, assumptions — across long conversations as a tree, not a transcript. Auto-resumes in any new session. Cuts context tokens by ~57% on real sessions.

Lightweight session memory for Claude Code — one Stop hook, bring-your-own LLM (OpenAI/Anthropic/Ollama/DeepSeek), markdown+grep storage, explicit /sess recall, plus lifetime skill & MCP usage analytics (/ccskill, /ccmcp)

700 lines, zero dependencies. Persistent SQLite memory for Codex CLI. No server, no vector DB, no cloud — just a single Python file that ships with every Python installation.

A lightweight ChatGPT Projects protocol that keeps long idea-shaping sessions structurally coherent: GPT follows real gates and source boundaries, uses append-only internal notes to avoid drift, and produces clean final artifacts only after approval.

Local-first, lossless session records for AI coding agents.

Claude Code forgets everything between sessions. This plugin gives it persistent memory — automatic observation capture and context injection via Neuroloom.

An LLM-powered agent with calculator, web search, and sandboxed code execution tools — FastAPI + Ollama, SSE streaming with live reasoning traces, and Redis-backed multi-turn session memory.

Structured session memory for Claude Code — a Stop hook that keeps a 10-section summary.md updated every turn, so /compact and --resume don't lose your context.

Production-grade MCP composition pattern: four specialized servers orchestrated by LangGraph with config-driven notification routing, health checks, auto-rollback, Redis session memory, structured logging, and Bearer token auth. Mount pattern for local dev, distributed ports for production.

MCP server for AI agents: intelligent code exploration (AST-based symbol search, architectural blueprints, call graphs, 90%+ token reduction) + persistent session memory (global preferences, project decisions)

Zero-dependency session recall for Claude Code — instant memory across compactions, ~50 tokens per prompt