Skip to content

Get automated code review

This guide reviews a proposed change automatically: matching profiles trigger focused review agents, and you accept or reject each finding. It assumes changes flow through reviewable Patches; if there is no Patch yet, create one first.

Before you start

You need a checkout where sase runs and a Patch with at least one commit in a reviewable status. Reviews run only once the Patch is ready for review and its checks have finished.

1. Add a mentor profile

Declare when review should run in the configuration file. Each profile names the files it matches and the reviewers it starts:

mentor_profiles:
  - profile_name: python_review
    file_globs:
      - "*.py"
    mentors:
      - mentor_name: style_checker
        role: "Python style expert"
        focus_areas:
          - focus_name: style
            description: "Style and naming conventions"

Profiles defined next to the project apply to that project only; profiles in the home configuration apply everywhere unless restricted. See the mentors reference for match criteria and scoping.

2. Confirm the profile matches

Trace which profiles match the Patch and why before waiting for a review:

sase config mentor-match my_feature

The report shows each criterion result. Adjust the globs or diff patterns until the intended profile matches, then move the Patch to a reviewable status so the review agents start when capacity allows.

3. Review the findings and apply what you accept

Open the review modal for the Patch with ,C, move between findings, and press Space on the ones to keep. Then press a to propose the accepted comments as an amend for further review, or A to commit them directly. Review agents appear under the review tribe in the agents view, where they can be inspected or stopped like any run.

If it goes wrong

No review with a matching profile usually means the Patch is not reviewable yet or its checks have not finished: confirm the status and wait for the checks. For a review agent that never starts, see Unstick a queued agent.

  • Mentors — profiles, severities, output format, and every key in the review modal.
  • Automation and human checkpoints — where automated review fits among autonomous outcomes and human checkpoints.

_[project]: A sase project is a named unit of work registered with SASE. _[workspace]: A sase workspace is a numbered clone of a project's primary repo, managed by the workspace store and tracked in that project's registry.json. _[agent]: A sase agent is an agent session or a single agent that does not belong to a session. _[prompt]: A prompt is the text SASE hands to a provider to start an agent turn. _[macro]: A macro is a reusable prompt triggered with #name in agent prompts. _[Patch]: A Patch is SASE's local unit of change. _[bead]: A bead is one tracked work item in SASE's git-native issue tracker. _[plan]: A plan is a persisted, approved artifact recording the intent behind agent work. _[TUI]: The TUI is SASE's terminal interface for running agents and reviewing results. _[epic]: An epic is an executable multi-phase plan run with sase bead work. _[gate]: A sase gate is a durable, command-backed request for one user decision, created when work must pause for the user. _[monitor]: A sase monitor is a monitor turn that runs one long command under a detached supervisor, so the command outlives the agent that started it. *[stitch]: A stitch is the lightweight ordered change record inside a Patch's STITCHES: section.