vllm-project/semantic-router

feature: Memory Assisted Code Review Bot

Open

#1,902 opened on May 11, 2026

View on GitHub
 (8 comments) (0 reactions) (1 assignee)Go (699 forks)github user discovery
area/agentarea/environmentarea/researcharea/testingarea/tool-managementarea/user-experienceenhancementgood first issuehelp wantedpriority/P1stale

Repository metrics

Stars
 (4,293 stars)
PR merge metrics
 (PR metrics pending)

Description

Describe the feature

Since we have many code agent authored PR, we need to have a more streamlined review process to help AI review bot to understand these PRs, especially the internal reasoning of these coding assistant. Most of these internal thinking are in the memory md files.

I checked GitHub marketplace, the existing AI review actions only review the diff. So I am proposing that:

  • customize AI review action
  • require submitting the memory md or revised form (make sure no privacy info is in the md file) in docs directory when submitting a large code change (e.g. >500 LOC)
  • instruct the AI review action to review the diff based on the memory md.
  • publish this customized review action in GitHub marketplace

Primary layer

global level

Why this layer?

Coding agents of different kinds make review process very challenging. At the same time, we are also facing GitHub Copilot review limit. Having a customized review action is the way to go.

Why do you need this feature?

No response

Additional context

No response

Contributor guide