Memory plugins

Give your AI persistent memory by mining projects and conversations into a searchable palace, enabling it to recall past work, decisions, and facts during future sessions.

Create and manage an autonomous AI agent company directly from chat — onboard organizations, design governance-aware agents, convert plans into parallelized tasks, monitor work, approve outputs, and coordinate agent-to-agent operations. Includes a persistent memory system using the PARA methodology for cross-session knowledge management.

Retrieve version-specific documentation and code examples from source repositories into LLM context using Upstash Context7 service.

Automatically stores and retrieves long-term memory for Claude Code sessions, recalling relevant context before each prompt and retaining conversation transcripts after responses.

Equip Claude with creative writing skills for fiction: character simulation, story planning, prose drafting, multi-level editing, research, and reader-response feedback. Build and staff writing teams with specialized agents for in-character roleplay, canon checking, style analysis, and structural outlining.

Persist and recall session memory across Claude Code sessions, automatically loading context on start and saving on stop, with skills to store and search shared memories via Bash API calls.

Maintain a persistent Zettelkasten knowledge graph of insights from your coding sessions — create atomic memory cards with bidirectional links, sync them across devices via git, load prior context when tasks overlap past work, and audit card health for orphans or contradictions.

Daily language tutoring with spaced repetition (SM-2) across reading, writing, speaking, and vocabulary drills, plus adaptive session planning and progress tracking via /fluent commands.

Enable AI assistants with persistent, relational memory across sessions using graph-vector storage. Store, recall, update, delete, and link memories for durable context awareness.

Maintain a persistent biography of the user across AI agent hosts (Claude Code, Codex, OpenClaw, Linggen) using a three-tier memory model, with session initialization and swarm routing for multi-agent distribution.

Persist and search coding session insights as atomic memories, surfacing past decisions and blocking resolutions for ongoing context

Query and traverse an Obsidian vault as a local knowledge graph — search semantically, find paths between nodes, detect communities, and analyze graph relationships using natural language claims.

Persist and recall working memory across Claude Code sessions — load relevant memories on start, capture decisions and outcomes at session end, search and inspect past context on demand, and automatically consolidate memories over time with decay-aware retention.

Continuously remembers context across Claude Code sessions via Honcho, eliminating the need to repeat preferences and project details. Includes interactive setup, preference surveys, and session lifecycle management with shell integration.

Persist and recall project knowledge, decisions, and constraints across Claude Code sessions using a local knowledge graph. Automatically captures insights during conversations and loads relevant context at session start, so your AI assistant never starts blind.

Persist and retrieve decisions, learnings, and context across Claude Code sessions, organizing knowledge into namespaced memories that are automatically surfaced when relevant via hybrid, vector, or text search, and analyze project artifacts for memory integration gaps.

Give Claude Code persistent memory across sessions, recalling past work, decisions, and error solutions via a structured 3-layer search workflow powered by a local Node.js memory tool.

Combine external context retrieval from Upstash's Context7 service with AI-assisted code generation and editing via Codex CLI, enabling memory-backed development and project management.

Persist and recall project knowledge, architectural decisions, and session context across Claude Code sessions using Cognis memory. Automatically index codebases to capture structure and patterns, enabling continuity between sessions.

Losslessly compress Claude Code conversation context using DAG-based summaries that preserve every message, allowing you to explore past discussions, promote insights across sessions, and diagnose memory pipeline health with compaction review and automated diagnostics.

Give Claude Code persistent, markdown-first memory across sessions with hybrid BM25 + dense retrieval and optional Haiku enrichment for automatic context recall.