SerialMemory for Developers

Persistent, queryable memory for AI agents. Built for engineers who want real long-term context.

MCP Protocol Native Zero Config Docker Full Tenant Isolation
Core Capabilities

What You Get

Long-term AI Memory

Store memories with semantic embeddings and full-text search. Memories persist across sessions, decay naturally over time, and can be reinforced when validated.

Graph Intelligence

Automatic entity extraction builds a knowledge graph from your memories. Multi-hop traversal lets you find connections your agent would otherwise miss.

Deterministic Reasoning

Not probabilistic guessing. Real constraint propagation, contradiction detection, and engineering validation over your system's actual structure.

Getting Started

How It Works

1

Create Account

Sign up at serialcoder.io. Free tier includes 10k memories.

2

Get API Key

Generate a key from your dashboard. One key per workspace.

3

Run Docker MCP

Pull and run. First launch prompts for your API key interactively.

4

Query Memory

Your AI agent now has persistent memory via MCP tools.

Zero Config

Docker MCP Server

One command to persistent AI memory.

terminal
docker run -it serialcoder/serialmemory-mcp

Interactive Setup

First run prompts for your API key

Config Stored

Credentials saved automatically in volume

No Env Vars

No manual configuration required

MCP Tools

Core API

Every tool your AI agent needs for persistent memory.

memory_ingest
memory_search
memory_multi_hop_search
memory_about_user
crawl_relationships
export_memories
get_graph_statistics
reembed_memories

Build agents that remember.

Give your AI a real memory layer. Semantic search, graph traversal, and deterministic reasoning—all through a single MCP server.