Scopes

A scope tells Kerno which endpoints to act on for a given operation.

The four scopes

Scope	Meaning
all	Every endpoint in the selected application
changed	Only endpoints affected by your current git changes (validation only)
file:path/to/handler.ts	Endpoints defined in that source file
endpoint:METHOD /path	A single endpoint, identified by HTTP method and path

When to use each

• all is useful for first-time baselining or when you want full coverage across an app.

• changed is the default for day-to-day validation. Kerno uses your git diff to figure out which endpoints to re-run, so you do not validate untouched code.

• file: scopes to a single source file when you have made focused changes to one handler.

• endpoint: validates one specific route when you want to verify a precise change.

Where scopes apply

These MCP tools take a scope parameter.

• kerno_list_endpoints — required

• kerno_plan_baseline — async mode only

• kerno_implement_baseline

• kerno_validate

The changed scope only works with kerno_validate.

How scopes work behind the scenes

Kerno's dependency graph maps every endpoint to the code that produces it. When you pick a scope, Kerno resolves it to a concrete list of endpoints, then plans, generates, or validates against that list. The mechanism is the same whether you drive Kerno through MCP or the IDE.

See How Kerno works for the validation flow, and Baselines and scenarios for what scenarios cover.