Skip to content

Selvedge vs. Git AI

Git AI stores attribution metadata in git notes and plugs into the Agent Trace alliance. Selvedge is complementary: it’s a compatible Agent Trace producer that adds live reasoning capture, a query layer, and changesets on top.

SelvedgeGit AI
Reasoning sourceCaptured live, by the agent in the same context that produced the changeAttribution metadata
GranularityEntity — DB column, table, env var, dep, API route, functionLine
MechanismMCP server — the agent calls it as work happensGit hook + Agent Trace alliance
GroupingChangesets — named feature/task slugs across many entitiesNone
Prior attemptsYesprior_attempts surfaces tried-and-rejected pathsNone
StorageSQLite, zero depsGit notes
  • Reasoning, not just attribution. Git AI records who/what touched a line. Selvedge records why, in the agent’s own words, captured live.
  • A query layer. selvedge blame, diff, history, changeset, and search are first-class. Git-notes attribution has no equivalent CLI you can pipe.
  • Agent Trace interop, not rivalry. Selvedge emits Agent Trace records too — see the Agent Trace interop page. The two layers compose: Git AI / Selvedge produce, downstream tools consume.
  • Entity-level + changesets for querying the history of a column or a whole feature, not a line.

If all you need is line-level attribution metadata living in git notes and surfaced through the Agent Trace ecosystem — and you don’t need captured reasoning or a query layer — Git AI is a focused fit.