servers/ memory/ spm
SP

SPM

Unverified Claim & verify → Memory

Agent memory that refuses to guess: evidence-gated recall, exact-source reads, verifiable deletion.

Visit website ↗ Try Remote Server
Overview Installation

Overview

SPM is a Model Context Protocol server that provides agent memory built on evidence-gated recall, exact-source reads, and verifiable deletion. Any MCP-compatible AI assistant can call SPM's tools directly, letting the model decide when to use them during conversations or agent runs.

SPM runs as a remote hosted service at https://api.spmos.ai/mcp, requiring only an API key for authentication. There's nothing to install locally—connect your client to the endpoint and start using it immediately. The latest release is v3.0.0.

SPM is listed in the Memory category on mcp.site and works with MCP clients including Claude Desktop, Claude Code, Cursor, VS Code, Windsurf, Zed, and others. If you maintain SPM for spmos.ai, claim this listing to verify ownership and keep the information current.

▸Hosted endpoint — Streamable HTTP · https://api.spmos.ai/mcp

Installation

Settings → Connectors → Add custom connector
Name: SPM
URL:  https://api.spmos.ai/mcp

Related servers

PM
PLUR Memory
Memory

Persistent long-term memory graph shared across MCP clients.

★ 226 Added 1 week ago →
KI
Kin
Memory

Personal relationship memory for assistants that remember people.

★ 45 Added 3 days ago →
NoteMCP icon
NoteMCP
Memory Remote

Give Claude and ChatGPT long-term memory. Save notes by text, voice or share sheet, and your AI finds and updates them in any future chat.

Added today →
Mcp icon
Mcp
Memory Remote

AI memory layer — one shared, persistent memory across every AI tool you connect.

Added 1 month ago →
BO
Bourdon
Memory

Recognition-first cross-agent memory federation. One shared memory across all your agents.

Added 1 month ago →
MC
Mcp
Memory Remote

Persistent memory for AI assistants. Save once; recall from Claude, ChatGPT, or any MCP client.

Added 1 month ago →
>_ esc
↑↓ navigate↵ openesc close