Empirical

Name: Empirical
Author: mattgauzza

by mattgauzzaversion 1.0.1i★ –iRuns in the cloud

Registry nameio.github.mattgauzza/empirical

Install

Overview

Description from the maker (original English)

Persistent memory for AI agents: save, search and connect context across tools and sessions.

PopNx MCP’s three pillars

These three are never merged into one score: opinions, measurements and popularity are different things.

User voiceiNo reviews yetReviews open together with accounts in a later step
Our basic checksi3 of 3 passedChecked from public information; code not scanned yet
UsageiNot measured yet★ –

What you can do with it

We have not written a plain-language summary for this one yet. Below are the facts we read from the registry.

Connection: Runs in the cloud
Transport: Streamable HTTP
Category: Search & AI memory

Install

Pick the AI app you use, then copy the settings.

Before you install

Basic checks: 3 of 3 passed
Runs on the maker’s server: what you ask it goes through their server.
The registry lists no key to fill in.

You need

Runs in the cloudNothing to install; it connects over the internet

Which app do you use?

ClaudeClaude CodeCursorVS CodeChatGPT

Pick your app above. We remember it, so next time the button says “Install in …”.

Works in Claude on the web, desktop and mobile:

Open Settings → Connectors
Press Add custom connector
Paste the address below and press Add
If it asks you to sign in, sign in with the maker

https://empirical.gauzza.com/mcp

Run this in your terminal

claude mcp add --transport http empirical https://empirical.gauzza.com/mcp

Open in Cursor

Or copy the settings yourself

Paste into .cursor/mcp.json in your project, or ~/.cursor/mcp.json

{
  "mcpServers": {
    "empirical": {
      "url": "https://empirical.gauzza.com/mcp"
    }
  }
}

Your browser asks to open the app, then the app shows the settings and asks you to confirm. Key values are left empty for you to fill in.

Open in VS Code

Or copy the settings yourself

Paste into .vscode/mcp.json in your project

{
  "servers": {
    "empirical": {
      "type": "http",
      "url": "https://empirical.gauzza.com/mcp"
    }
  }
}

Your browser asks to open the app, then the app shows the settings and asks you to confirm. Key values are left empty for you to fill in.

ChatGPT connects only to servers on the internet:

Open ChatGPT on the web → Settings → Connectors (Apps)
Press Create or Add (some plans need Developer mode under Advanced first)
Paste the address below, give it a name and save
If it asks you to sign in, sign in with the maker

https://empirical.gauzza.com/mcp

Menu names can differ slightly by plan and over time.

Your choice stays on this device in a cookie; no account needed.

These settings are generated from the maker’s registry entry. If the README says otherwise, follow the README.

What it can do

The server asks you to sign in before it lists its tools.

Basic checksi

We check information anyone can see; we have not read the code. Checks that do not apply are left out of the count. · Checked Nov 7, 2026

✓ Pass
Maker matches the repo
The registry name was verified with the GitHub account mattgauzza, which owns the repo
… Not checked yet
Still maintained
We have not read this repo from GitHub yet
… Not checked yet
Has a licence
We have not read this repo from GitHub yet
✓ Pass
Endpoint responds
Responded in 141 ms (HTTP 401)
✓ Pass
Sign-in with OAuth
Declares OAuth sign-in, so your AI app can take you to log in
… Not checked yet
No hidden instructions in tool descriptions
Not checked yet; we read the tool list when we check the endpoint

Compared with others in Search & AI memoryi

More stars than 0% of the other 7251 servers in this category
Published to the registry more recently than 94% of the other 7251 servers in this category
Basic checks equal to or better than 100% of the other 7251 servers in this category

Compared with every active server in this category in the registry. Stars count only when the publisher owns the repo.

Common questions

Is Empirical safe?

It passed 3 of 3 applicable basic checks. We have not read the code. Before adding a key or token, see what access the server asks for and use a key with only the permissions you need.

What keys or codes does it need?

The registry does not list a required key, though some servers ask you to sign in on first use.

Does it run on my computer or in the cloud?

It runs in the maker’s cloud. Connect over the internet with nothing to install, but what you send passes through the maker’s servers.

Is it still maintained?

The GitHub repo last changed –, and the latest version (1.0.1) reached the registry 2 days ago.

Which apps can use it?

Apps that support MCP, such as Claude Desktop, Claude Code, Cursor and VS Code. It connects by URL, so apps that accept remote MCP servers work too, such as ChatGPT in developer mode.

What can I use instead?

In the same category: Context7, Codebase Memory, Exa

Alternatives

Servers in the same category from other makers, sorted by stars.

Reading a card:★ GitHub starsiThree pillars: reviews · checks · usageiUpdated = latest version published to the registryi

Context7

by upstash

Feeds AI the latest docs for libraries and frameworks before it answers, so code it writes does not use outdated, removed functions.

Computer or cloud★ 64,224MITUpdated today

Install

Codebase Memory

by DeusData

Codebase knowledge graph for AI agents — 162 languages, sub-ms queries, 99% fewer tokens.

Runs on your computer★ 48,148MITUpdated 20 days ago

Install

Exa

by exa-labs

Connect AI agents to Exa for web search, content fetching, and multi-step research.

Runs in the cloud★ 5,213MITUpdated last month

Install

README from the maker

Open the README

Copied from the maker’s repo as written. PopNx MCP has not reviewed or endorsed it. HTML, images and scripts are removed. Links lead off this site.

empirical-memory

Cross-tool, cross-project memory for Claude Code, Codex CLI, and GitHub Copilot CLI, powered by Empirical.

Start an idea in ChatGPT, pick it up in your coding agent with full context carried over. Track leads and relationships as durable memory instead of a separate CRM. Organize memory into separate workspaces for personal, team, or client contexts. Recalls prior decisions, preferences, and conventions across sessions and projects, and persists new durable ones as you work — so you don't have to re-explain context you've already given elsewhere.

What this marketplace provides

empirical-memory: the required core memory Skill and MCP server connection.
empirical-work-history: optional concise completed-work summaries.
empirical-scar-tissue: optional recurring failure patterns and prevention guidance.
empirical-todo: optional project-based todos with status and next actions.
empirical-voice: optional writing in your own saved Empirical writing voice (needs a voice saved with the paid Writing Voice interview).

Each capability is a separate marketplace plugin, so users can install or remove optional capabilities without changing the core memory connection.

Package documentation:

Core Empirical Memory
Work History
Scar Tissue
Todo
Voice

Install

Claude Code

claude plugin marketplace add mattgauzza/empirical-memory-plugin

claude plugin install empirical-memory --scope user claude plugin install empirical-voice@empirical-memory --scope user # optional, writing voice

Codex CLI

codex marketplace add mattgauzza/empirical-memory-plugin

codex plugin add empirical-memory@empirical-memory

or browse/install via the /plugins command.

GitHub Copilot CLI

copilot plugin marketplace add mattgauzza/empirical-memory-plugin

copilot plugin marketplace update empirical-memory copilot plugin install empirical-memory@empirical-memory

or use the /plugin install slash command.

Recommended: install the latest `empirical` CLI

The MCP tools work standalone, but the Skills prefer the empirical CLI when it is available. The CLI provides the most reliable auth path, direct memory commands, and avoids an extra MCP tool-discovery round trip. It is not bundled with this plugin; install or update the latest version separately:

npm install -g @empirical-gauzza/cli

For a first-time setup, run the interactive installer to connect the CLI and supported coding agents:

empirical install

Installing or updating the latest CLI does not require rerunning empirical install. Use empirical plugin install all when you only want to refresh the native plugin, or let the CLI offer a one-time update when a newer marketplace plugin is detected during interactive use. You can check immediately with empirical plugin check --force.

For normal interactive CLI use without a permanent API key, run empirical auth login to sign in through the browser and save a local OAuth session. Headless bootstrap is still available for automation and external runtimes.

If you only want the plugin's MCP connection, you can skip the CLI installation.

If the CLI is already installed and you only need to install or refresh the native plugin, use:

empirical plugin install codex       # or copilot, claude, or all

Add optional capabilities explicitly with --work-history, --scar-tissue, --todo, --voice, or --all-optional. For example, empirical plugin install claude --voice installs empirical-voice, which writes in your saved Empirical writing voice (a paid feature; it needs a voice saved first). You can also use the native marketplace command: claude plugin install empirical-voice@empirical-memory --scope user.

This does not run the full MCP/auth questionnaire or modify global instruction files.

Optional capabilities selected by the empirical install questionnaire are installed through the same marketplace pathway. Neither records raw transcripts or secrets. Personalized answers are stored as the server-side Empirical policy and cached locally for native skills; they are not written into global AGENTS.md, CLAUDE.md, or Copilot instruction files.

Mode-aware skill content

skills/empirical-memory/ contains four files, not one:

SKILL.mcp.md — for installs where the user chose MCP-only. Leads with MCP tools, mentions the CLI only as an aside.
SKILL.cli.md — for installs where the user chose CLI-direct. Leads with empirical CLI commands as the confidently-recommended path, MCP mentioned only as a fallback.
SKILL.both.md — for installs where the user chose both. CLI preferred where available (a latency/scripting preference, not a safety one now that both paths dedupe and auto-link the same way), MCP as fallback.
SKILL.md — the file every host's native plugin mechanism actually reads (Claude Code, Codex CLI, and GitHub Copilot CLI all resolve a plugin's skill from the literal SKILL.md filename; none of their manifest schemas support selecting among sibling variant files at install time). Kept identical to SKILL.both.md, since "both" is the safest default for any consumer that hasn't been made mode-aware (an old CLI version, a direct git clone, a host with no CLI installed at all).

The empirical CLI is what actually makes the mode choice matter: after a native plugin install or update, it overwrites the on-disk SKILL.md for that host with the variant matching the user's chosen integration mode (see packages/cli/lib/installers/nativePlugin.js in the main empirical_v2 repo). If that overwrite step is missing, out of date, or fails, every install still gets a working "both" skill from this repo — never a missing or broken one.

When editing skill content: update SKILL.mcp.md, SKILL.cli.md, and SKILL.both.md for the mode-specific "how to query" / "how to record" mechanics; keep the shared policy sections (Before doing work, Follow the current user policy, Persist durable context, Authored Skills and Kits, Authentication and failures) consistent across all three by hand — there is no build step that enforces this. After editing SKILL.both.md, copy it over SKILL.md so the two never drift:

cp skills/empirical-memory/SKILL.both.md skills/empirical-memory/SKILL.md

Maintaining the manifests

This repo publishes to four hosts, each with its own manifest format. All of them are generated from a single source of truth, meta/manifests.json:

Generated file	Host
`.claude-plugin/plugin.json`, `.claude-plugin/marketplace.json`	Claude Code
`.codex-plugin/plugin.json`	Codex CLI
`.github/plugin/marketplace.json`	GitHub Copilot CLI
`.agents/plugins/marketplace.json`	generic AGENTS.md

plus a .claude-plugin/ and .codex-plugin/ pair under each directory in plugins/.

Do not hand-edit those files. Edit meta/manifests.json, then regenerate:

node scripts/build-manifests.mjs           # write the manifests

node scripts/build-manifests.mjs --check # verify nothing drifted (CI runs this) claude plugin validate . --strict # same validator the plugin directory runs npm install --no-save yaml && node scripts/check-skills.mjs # every SKILL.md frontmatter must be valid YAML

To bump a version, change it in one place: the plugin's version in meta/manifests.json. The marketplace metadata.version follows the root plugin automatically.

Host-specific fields belong to their host. interface is a Codex/Copilot block; Claude Code silently ignores unknown top-level fields, so putting interface in a .claude-plugin/ file does nothing. Claude's equivalents are displayName, description, and homepage, and it has no counterpart for longDescription, category, capabilities, or defaultPrompt. The generator encodes this mapping.

License

MIT

Empirical

Overview

PopNx MCP’s three pillars

What you can do with it

Install

Before you install

What it can do

Basic checksi

Maker matches the repo

Still maintained

Has a licence

Endpoint responds

Sign-in with OAuth

No hidden instructions in tool descriptions

Compared with others in Search & AI memoryi

Common questions

Alternatives

Context7

Codebase Memory

Exa

README from the maker

empirical-memory

What this marketplace provides

Install

Recommended: install the latest empirical CLI

Mode-aware skill content

Maintaining the manifests

License

Recommended: install the latest `empirical` CLI