Provides local-first repository memory for coding agents to save and retrieve failed test results and project decisions.
Also known as OnurByte/Vexen, @onurbyte/vexen
- Stars
- 3
- Tools
- —
- Downloads, 30 days
- —
- Latest version
- —
- Last commit
- Nov 5, 2026
- License
- GPL-3.0
What is Vexen MCP?
Vexen provides local-first repository memory for coding agents to save and retrieve failed test results and project decisions.
Summary generated from the project's README
About Vexen MCP server
Coding agents that need to remember past mistakes can use Vexen to maintain a local-first repository memory. Instead of relying on long chat histories, this tool allows an agent to save short notes regarding failed test results and specific project decisions directly within a project folder. This ensures that even if you switch from Claude Code to Cursor or Codex, the lessons learned from previous failed attempts remain available to the next session.
Setup is performed locally inside a Git project using the command npx @onurbyte/vexen setup, which creates a .vexen folder to store data. The system is designed to be private, operating entirely on your machine without cloud storage, accounts, or telemetry. It automatically manages context by stepping aside if the relevant files change, preventing outdated information from influencing new code. This local approach works across various environments including macOS, Linux, and Windows.
What you can do with it
- 1Preventing repetitive failed fixesAsk your agent to check for previous failed attempts in a file before it tries to apply a new fix.
- 2Maintaining cross-tool project memoryUse the saved lessons from a session in one coding tool to inform a different tool in a later session.
- 3Recording specific test failuresInstruct your AI to save the exact test result and file location when a coding task fails.
Written from the project's README. Every use case is checked against a quote from it.
How to install Vexen MCP
Formats verified 10 Nov 2026Add to CursorAdd to GooseInstall Vexen MCP in Claude Code
$ claude mcp add vexen -- npx -y @onurbyte/vexenInstall Vexen MCP in Cursor
~/.cursor/mcp.json
{
"mcpServers": {
"vexen": {
"command": "npx",
"args": [
"-y",
"@onurbyte/vexen"
]
}
}
}Install Vexen MCP in VS Code
.vscode/mcp.json
{
"servers": {
"vexen": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@onurbyte/vexen"
]
}
}
}Install Vexen MCP in VS Code CLI
$ code --add-mcp '{"name":"vexen","type":"stdio","command":"npx","args":["-y","@onurbyte/vexen"]}'Install Vexen MCP in Claude Desktop
claude_desktop_config.json
{
"mcpServers": {
"vexen": {
"command": "npx",
"args": [
"-y",
"@onurbyte/vexen"
]
}
}
}Install Vexen MCP in Codex
$ codex mcp add vexen -- npx -y @onurbyte/vexenInstall Vexen MCP in Codex config.toml
~/.codex/config.toml
[mcp_servers.vexen] command = "npx" args = ["-y", "@onurbyte/vexen"]
Install Vexen MCP in Gemini CLI
$ gemini mcp add -s user vexen npx -- -y @onurbyte/vexenInstall Vexen MCP in Gemini settings.json
~/.gemini/settings.json
{
"mcpServers": {
"vexen": {
"command": "npx",
"args": [
"-y",
"@onurbyte/vexen"
]
}
}
}Automated checks
Ran Nov 14, 2026Full report →Passed our automated checks. That is not a security audit: read the code and give the server the narrowest access it needs. How checks work
- Source repository linked and reachableonurbyte/vexen is public and reachablegithub.com/onurbyte/vexen ↗Pass
- Package names this repository as its source@onurbyte/vexen names onurbyte/vexen as sourcenpmjs.com/package/@onurbyte/vexen ↗Pass
- Build provenance (Sigstore attestation)Sigstore attestation for @onurbyte/vexennpmjs.com/package/@onurbyte/vexen/v/1.0.0#provenance ↗Pass
- Open source license (OSI)GPL-3.0spdx.org/licenses/GPL-3.0.html ↗Pass
- Commit or release in the last yearLast commit or release 2026-11-05github.com/onurbyte/vexen ↗Pass
- OpenSSF ScorecardNot checked
- Package name not close to a popular packageNo popular package one edit away from @onurbyte/vexenPass
- No critical advisory in the package or its dependenciesNot checked
- No hidden instructions to the model0 tool descriptions and the README read; no override phrases, credential paths or hidden charactersPass
- Destructive tools are labelledNot checked
- Remote endpoint answers the MCP handshakeNot checked
README
View on GitHubnpx @onurbyte/vexen setup
The problem, and the fix
You don't need to save a conversation. Your agent saves a short note: what failed, in which file, with its test result. A later session checks it before editing the same code. If the file changes, the note steps aside.
The evidence
A solved task cost 77.7% less with Vexen memory in our pilot: 4.5× cheaper.
The saving comes from more runs succeeding, not from shorter runs. In the first published pilot, Claude Code with Haiku solved 55% of tasks with relevant memory, against 10% without (11 of 20 against 2 of 20).
These are small, honest numbers: one model, synthetic tasks, notes prepared in advance. They do not show general coding improvement. Method, raw data, stale-note and control checks, and limits →
Install
Run it inside the Git project you want Vexen to remember. You need Node.js 18 or newer (only for this installer) and Git.
On Windows PowerShell use npx.cmd @onurbyte/vexen setup. setup creates your project's memory (a local .vexen/ folder), keeps a permanent copy of Vexen on your machine, connects the coding tools it finds, then starts Vexen once and checks that it answers. It prints what it did and what it could not do. Nothing runs in the background.
Then: restart your coding tool and ask:
Use Vexen to check this project's README.md for past failures or decisions.
On a new project, “no relevant evidence” is the expected answer. When a test fails during your work, ask your agent to save the failed attempt with its file and test result. If your agent can't find Vexen, run npx @onurbyte/vexen doctor --mcp.
Full setup guide, manual steps and uninstall → · Download without Node.js →
Works across coding tools
Start a task in Claude Code. Continue in Codex or Cursor. Saved lessons stay with the project.
The installer supports eight tools; more have completed real Vexen tool calls. Setup and coverage vary. See what was tested →
Your project. Your memory.
- Local storage. Notes live in your project's
.vexen/folder. Database files are ignored by Git. - No Vexen cloud or telemetry. Vexen doesn't record your chats. Your coding tool's own data policies still apply.
- A source for each reminder. See the file, commit and recorded result behind a note.
- Nothing relevant? It says so. Old or unrelated evidence should not become advice for new code.
Version 1.0: the command line, the four MCP tools and the data format are stable and follow semantic versioning. Vexen gives your agent context to consider; the agent still needs to inspect the current code and test its work. What 1.0 does not claim is in the roadmap.
Privacy · Documentation · Compatibility · Roadmap · Report an issue
GPL-3.0-or-later · License. Independent; not affiliated with Vercel. Geist fonts are used under the SIL Open Font License. Product and platform marks identify compatibility and are not endorsements.
Vexen MCP questions
How do I add Vexen MCP to Claude Code?
Run claude mcp add vexen -- npx -y @onurbyte/vexen in your terminal. Add --scope user to make it available in every project.
How do I add Vexen MCP to Cursor?
Open ~/.cursor/mcp.json and add the server under mcpServers, using the Cursor config shown in the install section above.
How do I install Vexen?
You can install it by running npx @onurbyte/vexen setup inside your Git project. Node.js 18 or newer is required for the installer.
Where is the data stored?
All notes and memory are stored locally within your project's .vexen/ folder.
Does Vexen use the cloud or collect telemetry?
No, Vexen is local-first with no cloud storage or telemetry, ensuring your data stays on your machine.
Which coding tools are compatible?
Vexen works across various tools including Claude Code, Codex, Cursor, Gemini CLI, and VS Code.