MCP servers that give assistants persistent memory, knowledge graphs and structured reasoning. Each entry is scanned with the deterministic Capability-Flow Trust Model — grades are computed, never self-reported.
Session management for TMCP (TypeScript Model Context Protocol) transport implementations. This package provides the base classes and in-memory implementations for both streaming session coordination and session metadata persistence.
The official structured-reasoning server: break problems into revisable thought sequences.
The official knowledge-graph memory server: persistent entities, relations and observations.
MCP Apps UI resources and server helpers for n8n
Open-source MCP server exposing Agent402.Tools' catalog — 500+ strong: 400+ self-hostable tools + 100 multi-tool skill packs (security-audit, trend-analysis, structured-scrape, decode-blob, forecasting-bake-off) for AI agents — browser, web search & answe
Modern MCP server for Ollama – rebooted and actively maintained.
MCP server providing comprehensive access to BookStack knowledge management system
MCP server that gives coding agents persistent project understanding via a knowledge graph
Headless semantic MCP server for Obsidian, Logseq, Dendron and any markdown-based knowledge base
A Model Context Protocol server for knowledge graphs designed around neurodivergent thinking patterns
Neo4j-based knowledge graph with temporal versioning and semantic search.
MCP server for little big brain — graph and hybrid search tools for agents
MCP server for XMind
Model Context Protocol server for FoundryVTT integration
MCP server for local markdown notes — ranked search, tags, todos, and a wiki-link knowledge graph, from any MCP client.
Read-only MCP server over robertdelanghe.dev's signed static API. A thin implementation of @bounded-systems/static-mcp: the site's verbs + resource catalog + config, served as a Sigstore-verified static-response MCP server over stdio.
MCP server for Mushi Mushi — triage user-felt bugs, pull evidence (screenshots, console logs, network requests), check blast radius, and dispatch fix PRs, all from your AI coding agent
Persistent context management for AI coding assistants so context survives across sessions.
MCP server for cosense
MCP server for SERPdive, the AI Search API: real-time web search returning answer-ready page content for LLMs and agents. Same speed as Tavily, 20.2% fewer tokens, 60.7% of decided quality duels on a public benchmark.
Zero-dependency MCP server that gives AI agents a self-updating project memory in AGENTS.md. Returns merge instructions instead of mutating state, so every change is a reviewable diff.
Safe, structured Model Context Protocol server for SiYuan Note
BuilderForce Agent Memory — MCP transport. Exposes @seanhogg/builderforce-memory to any MCP client (Claude Agent SDK in-process, stdio, or HTTP) over a pluggable MemoryBackend, with token-saving fetch-on-demand recall.
MCP Server for SlimContext - AI chat history compression tools
Servers in this category extend an AI assistant with state that outlives a single chat. Instead of forgetting everything when the session ends, the assistant writes to and reads from an external store through the Model Context Protocol. Typical backends include vector databases such as Chroma or Qdrant, graph stores like Neo4j, embedded SQLite, plain JSON or Markdown files on disk, and hosted memory layers such as Mem0 or Zep. A second group adds no storage at all: structured-reasoning servers give the model a scratchpad for step-by-step thinking, branching and revision. The full MCP Trust Registry lets you compare them side by side.
Memory pairs naturally with servers from the Databases and Productivity & Workflow categories, where the recalled context drives real queries and actions.
A memory store is a write path into every future conversation. If a malicious or compromised MCP server — or untrusted content the assistant summarizes into memory — plants a poisoned entry, that entry is replayed as trusted context indefinitely: persistent prompt injection. Memory files also accumulate exactly what attackers want, from internal plans to credentials mentioned in passing, and a server that combines memory with broad file or network access could quietly exfiltrate that entire history.
That is why each listing here carries a Trust Score: a deterministic engine reads the published npm or PyPI package, maps its capability blast radius and reports individual findings with evidence. The A-F grade is reproducible, involves no AI judgment and cannot be bought — the scoring model is documented in full.
It can be, when the memory MCP server has a narrow blast radius — local storage, no unnecessary network egress — and a strong Trust Score. Prefer servers whose stored data you can inspect and wipe. Treat everything written into memory as future prompt input, because that is exactly what it becomes.