Selvedge runs on your own machine — the client starts it, so there is no endpoint to ping. 223 installs a week from pypi. Last commit 15 Sep 2026.
Decision provenance for AI-coded codebases: the why, and what was already tried and rejected.
We read the source, 12 h ago · rules 3dff92dd89df
What this server is able to do. For an MCP server this is often the job itself — a terminal server runs commands because that is what it is for. Listed so you know what you are plugging in, not as an accusation.
const child = spawn(cmd, args, { stdio: "inherit", shell: needsShell });
const candidate = path.join(dir, cmd + ext.toLowerCase());
result = subprocess.run(
Is this your server and something here is wrong? Tell us — corrections are free and do not require a plan.
We found places where it runs commands, builds paths or queries from values it is given. None of that is a flaw by itself — it becomes one when the code changes, and code changes quietly between releases. We re-read it on every one.
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add selvedge -- uvx selvedge
{
"mcpServers": {
"selvedge": {
"args": [
"selvedge"
],
"command": "uvx"
}
}
}
[mcp_servers.selvedge]
command = "uvx"
args = ["selvedge"]
{
"mcpServers": {
"selvedge": {
"args": [
"selvedge"
],
"command": "uvx"
}
}
}
{
"mcpServers": {
"selvedge": {
"args": [
"selvedge"
],
"command": "uvx"
}
}
}
This one needs environment variables set before it will start:
SELVEDGE_DB (Override the SQLite path. Defaults to .selvedge/selvedge.db next to the codebase.).
The author declared them in the registry entry; get the values from the project itself.
Cited 'why was this decided?' answers for your repo and for React, Kubernetes, and Rust.
Read-only visualization for Cortex memory, knowledge, sessions, traces, and codebase graphs.
Database-driven FP enforcement and project management for AI-maintained codebases
Ship, discover and review vibe-coded projects on Shipyard, the invite-only feed for AI-built apps.
Project memory across sessions for AI agents. A queryable decision log in the repo.
Finds real dead code and proposes a safe, test-gated branch+PR to remove it.
Pack local or remote codebases into AI-friendly files that LLMs and coding agents can read or search
Self-hosted session relay. Shared threads across Claude Code, Cowork, and Claude.ai.
Answers built from our own checks of this server.