#agent-memory

MCP server that exposes agent-memory-daemon to any MCP-compatible client — Kiro (CLI & IDE), Claude Desktop, Cursor, and others. The daemon does the thinking (consolidation + extraction); this server is a thin filesystem bridge so agents can read, append, and search memory through the Model Context Protocol.

scratchpad-mcp is an MCP server that gives AI agents persistent, token-efficient storage. It solves a specific waste problem: agents constantly re-read files they've already seen, re-summarize documents they've already processed, and re-load context they've already understood. Every one of those round-trips burns tokens for no new information. This server fixes that with eight tools designed around how agents actually work: Versioned writes. write_file automatically versions every write and keeps the 10 most recent versions per file. Storage is append-only on success and atomic on failure partial writes can't corrupt state. Structured diffs. read_file accepts a since_version parameter and returns a JSON line-diff against that prior version instead of the full content. Agents that have already seen v1 can ask "what changed in v3?" and get a small structured payload they can reason about, not the entire file again. Append-only logs. append_log and read_log give agents an event-stream they can replay. Cursor-based pagination (since_entry + last_entry_id + has_more) means an agent can checkpoint where it left off and resume cheaply. On-demand summaries. summarize_file calls Claude Haiku to summarize files over ~2000 estimated tokens. Summaries are cached per file version, so repeat calls on an unchanged file cost nothing. The threshold is enforced server-side you can't accidentally pay to summarize something small. Per-agent isolation. Every operation is scoped by an agent_id parameter, so one server instance can serve many agents without leaking state between them. Storage limits. 1 MB per file write, 64 KB per log entry, 1000 files / 100k log entries / 100 MB total per agent sane multi-tenant guardrails out of the box. Backed by a single SQLite file (Postgres migration is on the roadmap). All SQL is parameterized, paths are validated against a strict allowlist, and the security model is documented honestly it's safe for one-user-per-process deployments today, and the V2 plan derives agent_id from the caller's API key for true multi-tenancy. Build agents that remember what they've already seen.

pen-source MCP server that AI coding agents call as they work to log structured change events. Captures the entity (DB column, table, function, endpoint, dependency, env var), the diff, and the reasoning behind each change to a local SQLite database — preserving the intent that would otherwise evaporate when the agent session ends. Six MCP tools: log_change, diff, blame, history, changeset, and search. Answers questions like "when was users.stripe_customer_id added and why?" or "which agent added the payments table?" Zero-config (SQLite, WAL mode), Python 3.10+, MIT licensed. Like git blame, but for semantic entities and AI agents.

Memtrace — Structural Memory for AI Coding Agents The Problem Every AI coding agent — Claude Code, Cursor, Codex, Copilot — starts each turn completely blank. It re-reads raw source files and re-derives the full call graph, type hierarchy, and import tree from scratch on every single invocation. That structural rework burns 60–90% of the context window before any real reasoning begins. Less than 5% of tokens in a typical agentic coding session contribute genuine new intelligence. The rest is expensive, redundant noise — and it compounds: accuracy drops 40% as sessions grow, stale context crowds out signal, and summaries strip out the structural relationships agents need most. The Solution Memtrace is a bi-temporal structural memory layer that turns your codebase into a live, queryable knowledge graph — compiled from the AST, not guessed from embeddings. Every function, class, interface, and API endpoint becomes a typed node with deterministic relationships. Every file save becomes a queryable episode with timestamps, so agents can reason about structure, detect regressions, and time-travel through their own work without re-reading anything. One Rust binary. Zero configuration. Five-minute install. What agents can do with it Find callers, callees, and dependencies instantly — no file scanning, no token waste Compute blast radius before making a change — know exactly what breaks before anything is touched Detect structural drift between sessions — catch regressions the moment they happen, not at PR review Time-travel through code evolution — query any prior state of any symbol, not just git commits Search across the full codebase with hybrid retrieval — BM25 full-text + HNSW vector + graph traversal fused in one query Map API topology across services — cross-repo HTTP call graphs, dependency chains, dead endpoint detection Benefits −90% token cost on structural queries (Mem0) +26% accuracy on multi-step agentic tasks (Mem0) −91% p95 latency on structural lookups vs. RAG baselines +32.8% SWE-bench bug-fix success rate when agents have graph context (RepoGraph) 200–800ms per-save re-indexing — every file save is a queryable episode in under a second 40+ MCP tools covering indexing, search, relationships, impact analysis, temporal evolution, API topology, graph algorithms, and direct Cypher queries 12 languages + 3 IaC formats supported via Tree-sitter grammars Local-first, closed-source Rust — code never leaves the machine, no account required, no telemetry

Strata is a self hosted AI memory server. Your AI remembers everything across every session, on your own hardware. Key features: . Semantic search (find memories by meaning, not keywords) . Per-agent API keys with granular permissions . 3D constellation viewer with live agent activity . File vault (attach real documents to memories) . CSV audit log (full transparency on every agent action) . Pre-Strata history import (your memory doesn't start at install day) . Global MCP kill switch (emergency brake, only a human can undo) . Automatic deduplication . 10 structured thought types . Backend using PostgreSQL . Runs on a Raspberry Pi Always on, always local, always yours.

Persistent state and continuity MCP server for agents. Prepare context, record events, search prior memory, build task context, and inspect memory timeline over MCP-over-HTTP.

65+ AI tools as an MCP server. Research, write, code, scrape, translate, analyze, agent memory, workflows. Pay per call from $0.006.

18 MCP tools for AI agent infrastructure — persistent memory with FTS5, API drift detection with cron scheduling, and HMAC-signed trust receipts for audit trails

Git for agent memory. Memstate gives AI agents versioned, structured memory with automatic conflict detection, full version history, and up to 80% fewer tokens than vector-based alternatives. Extracts facts from text and markdown into structured keypaths, with 84.4% accuracy on the open-source benchmark vs 20.4% for Mem0.

Hypergraph database for AI agents. Store n-ary relationships (edges connecting 2+ nodes), track provenance (source and confidence), and query paths — all backed by local SQLite with zero configuration.

An MCP server that provides AI agents with a persistent, PostgreSQL-backed virtual filesystem. Supports session-isolated file operations, cross-session shared stores, 11 POSIX-style tools (read, write, ls, mkdir, rm, mv, glob, grep, etc.), and optional Row Level Security for multi-tenant deployments. Works with Claude Desktop, Claude Code, Cursor, Windsurf, Cline, and any MCP client.

Penfield gives AI agents long-term memory that compounds over time. Instead of starting every session from zero, your agent remembers conversations, learns preferences, connects ideas, and picks up exactly where it left off. Core capabilities: - 16 tools covering memory CRUD, knowledge graph, context management, artifacts, and reflection - Hybrid search with tunable weights — BM25 (keyword), vector (semantic), and graph traversal - Knowledge graph with 24 relationship types across 8 categories (evolution, evidence, hierarchy, causation, implementation, conversation, sequence, dependencies) - Context checkpoints for session handoff between agents or channels - Artifact storage for files, diagrams, and reference docs - 11 memory types intelligent filtering - Personality system with configurable personas via the portal Also available as a Claude Connector, Claude Code integration, native OpenClaw plugin (4-5x faster), and direct API.

Simple standalone MCP server giving Claude the ability to remember your conversations and learn from them over time.

The llm_to_mcp_integration_engine is a communication layer designed to enhance the reliability of interactions between LLMs and tools (like MCP servers or functions).

DevContext is a cutting-edge Model Context Protocol (MCP) server designed to provide developers with continuous, project-centric context awareness. Unlike traditional context systems, DevContext continuously learns from and adapts to your development patterns and delivers highly relevant context providing a deeper understanding of your codebase.

DevContext is a cutting-edge Model Context Protocol (MCP) server designed to provide developers with continuous, project-centric context awareness. Unlike traditional context systems, DevContext continuously learns from and adapts to your development patterns and delivers highly relevant context providing a deeper understanding of your codebase.

MCP server that execute applescript giving you full control of your Mac

AI agent that controls computer with OS-level tools, MCP compatible, works with any model

The Cursor10x MCP is a persistent multi-dimensional memory system for Cursor that enhances AI assistants with conversation context, project history, and code relationships across sessions.

A lightweight MCP server that integrates with Apple Notes to create a personal memory system for AI. Easily recall and save information from your Mac using simple AppleScript commands. Compatible with all macOS versions with minimal setup requirements.