What is Selvedge?

Selvedge is a local MCP server. AI coding agents (Claude Code, Cursor, Copilot) call it as they work to log structured change events with reasoning. Your data stays in a SQLite file under .selvedge/ next to your code.

The problem it solves

Human-written code leaks intent everywhere — commits, PRs, comments, Slack threads. AI-written code doesn’t: the agent’s perfect clarity about each decision lives in the prompt and evaporates when the conversation ends. Selvedge captures the why — live, by the agent itself, as the change is made. The diff is git’s job. The why is Selvedge’s.

What Selvedge captures

Each event records:

• What changed — entity path, change type, diff
• When — UTC timestamp
• Who — agent name, session ID
• Why — reasoning, captured from the agent’s context in the moment
• Where — git commit, project root

What “entity” means here

Most attribution tools work at the line level. Selvedge attributes things you actually search for:

users.email           DB column (table.column)
users                 DB table
src/auth.py::login    Function in a file (path::symbol)
src/auth.py           File
api/v1/users          API route
deps/stripe           Dependency
env/STRIPE_SECRET_KEY Environment variable

The first question after git blame is usually “what’s the history of this column?”, not “what’s the history of lines 40–48 of users.py?” — so Selvedge meets you there.

Prefix queries work everywhere. Searching users returns users, users.email, users.created_at, and any other entity under the users. namespace.

What “changeset” means here

A Stripe billing rollout touches the users table, two new env vars, three new API routes, one dependency, and four functions across the codebase. Tag every event with changeset:add-stripe-billing and you can pull the entire scope back later — even if the original PR was broken into eight smaller ones over a month.

Where it deliberately doesn’t go

Selvedge is not:

• A replacement for git — line-level what/when stays git’s job.
• A code review tool — review-time quality is a different problem.
• An LLM observability platform — call traces, token costs, model hops are tools like LangSmith and Helicone.
• A code-host AI assistant — GitHub Copilot’s PR summaries answer a different question.

It’s the provenance-as-first-class-citizen layer that those tools can reference.

How that’s different from “ask an LLM about the diff”

The category that’s emerging around this — sometimes called “git blame for AI agents” — mostly works by feeding the diff back to a second LLM after the fact and asking it to reconstruct intent. That’s better than nothing, but it’s a guess. The agent that made the change knew exactly why; by the time you ask a fresh LLM to explain the diff, that context is gone.

Selvedge takes the other approach: it gives the original agent a way to record its intent as it works. The reasoning is the agent’s own — written from the same context window that produced the change. No inference, no hallucinated explanations. And an empty reasoning field is itself a useful signal: the agent didn’t have one.

Next

Quickstart → Three commands. How it works → The MCP plumbing in plain prose. Comparison table → Selvedge vs. AgentDiff, Origin, Git AI, BlamePrompt.