Skip to main content
Glama
prest4u

With

by prest4u
OverviewSchemaRelated ServersScoreDiscussions
Python
Local

English · 中文

With.

Things that stay with you.

With is a persistent memory layer for your harness.

The source of truth is an owned SQLite file on your machine. People look at an Obsidian vault. Every tool talks to the same CLI or stdio MCP. Facts are invalidated, not deleted.

The product is With. The command is still eric-memory, so existing installs keep working.

Stay

A current fact is active. A retired fact is deprecated plus superseded_by. Default search does not treat history as current. Nothing is erased to make a new present.

That is the whole philosophy, in one line: leave what was true, and write what is true now.

Related MCP server: recollect

Architecture

flowchart TB
  quest[Install_or_daily_quest]
  cli[CLI_and_MCP]
  store[Owned_SQLite]
  vault[Obsidian_projection]
  harness[Registered_harness]
  quest --> cli
  harness --> cli
  cli --> store
  store --> vault

No app. No cloud. No login daemon. No admin rights. Zero third-party runtime packages. Python 3.10+ is enough.

Two tiers

Tier

Who

How

Simple

Anyone installing on their own machine

Paste quests/en/install.md into the AI you already use. Then open the Obsidian home note. Run the daily quest when you want the vault refreshed.

Full

Maintainers

The same, plus a read-only Holograph import, tests, and adapters.

Guides: Introduction · Simple · Full · CLI and MCP · Windows / Linux

Chinese copies live next to them in docs/ and quests/.

Install

python3 -m unittest discover -s tests -q
python3 scripts/repo_check.py
python3 bin/eric-memory --version
python3 bin/eric-memory --data-dir "$HOME/eric-memory-data" init --tier simple
python3 bin/eric-memory --data-dir "$HOME/eric-memory-data" status

On Windows use py -3 and %USERPROFILE%\eric-memory-data. Expand ~ once at init. After that, only absolute paths.

Or skip the terminal: give quests/en/install.md to your harness and answer the questionnaire.

The data directory never enters Git.

Official surface

Action

CLI

MCP

Initialize

init

(once, from the install quest)

Status

status

memory_status

Write

add

memory_add

Search

search

memory_search

Invalidate

deprecate

memory_deprecate

Sync

sync

memory_sync

Import

import holograph --source …

memory_import

Index a folder

index-files

memory_index_files

Register a tool

harness add

memory_harness_add

List registered tools

harness list

memory_harness_list

Do not INSERT into memory.db by hand. The contract is skills/严格技能.md. A Cursor template is .cursor/mcp.json.example.

What v1 will not do

  • Use Mem0, Graphiti, Cognee, or Supermemory as the kernel

  • Treat Obsidian as the source of truth

  • Scan the whole disk in silence

  • Treat a vendor leaderboard as acceptance

  • Ship a client zip, a cloud, or a background service

License

MIT

This server cannot be installed

RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd
A
license - permissive license
Not graded
quality - not tested
C
maintenance

How are these scores calculated?

Maintenance

–Maintainers
–Response time
–Release cycle
–Releases (12mo)
Commit activity

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Servers

  • engram

    TushaeBXN
    A
    license
    Not graded
    quality
    C
    maintenance
    Provides persistent, local-first AI memory across sessions via MCP tools for storing, searching, and retrieving context from past interactions.
    1
    MIT

  • recollect

    danbaileyza
    A
    license
    Not graded
    quality
    B
    maintenance
    Local, cross-agent memory for AI coding agents using a single SQLite file, enabling persistent sessions and durable facts shared across multiple MCP-compatible tools.
    14
    MIT

  • nuzo-memory

    fabionfsc
    A
    license
    Not graded
    quality
    F
    maintenance
    Local-first, auditable memory for AI agents. Provides durable context for MCP hosts with SQLite storage, CLI, and MCP tools for memory management.
    2
    Apache 2.0

  • Cairn

    QuarkOS
    A
    license
    Not graded
    quality
    B
    maintenance
    Enables agents to assert and recall typed facts with idempotent writes, freshness, and assurance verdicts. Provides MCP tools for persistent memory across sessions via SQLite-backed storage.
    724
    MIT

View all related MCP servers

Related MCP Connectors

  • XMemo

    Shared, governed long-term memory for AI agents across tools and sessions via MCP and REST.

  • ContextLattice

    Private-by-default, local-first memory/context/task orchestrator for MCP apps and agents.

  • mcp

    Your memory, everywhere AI goes. Build knowledge once, access it via MCP anywhere.

View all MCP Connectors

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/prest4u/with-memory'

If you have feedback or need assistance with the MCP directory API, please join our Discord server