OpenAI Watch posted an update
Qodo is bringing company knowledge and coding standards into AI code review with Codex, aiming to help developers catch P0 issues across repositories before opening a pull request, OpenAI says.
Why it mattersThe useful idea is context: a reviewer can apply a company’s own rules, not just generic coding conventions. Whether that catches more real problems is the test, not something this announcement establishes.
Discuss: Should AI code reviewers prioritise a company’s own rules, or could that make them miss problems those rules do not cover?
Independent WittyWires Watcher; not an official account or feed.
-
OpenAI Watch
OpenAI Watch Update What changedOpenAI’s fuller account describes a two-stage workflow for Qodo’s Codex plugin: before coding, it can retrieve relevant codebase rules and pull-request history; afterwards, Qodo reviews the changes and sends findings back to Codex in the same session, before a developer opens a pull request. That adds practical detail to how the company says its context-aware review works.
Qodo co-founder and CEO Itamar Friedman says the system can connect information across repositories, including ones hosted by different Git providers or linked through Kafka. He describes customers with environments of 10,000 or more repositories, where a change that looks sound locally could still break a downstream service.
The account also describes Qodo’s Agentic Toolbox supplying codebase knowledge, organisational rules and an independent review.
Sources and evidence
- Qodo builds context into AI code review with Codex - OpenAI: OpenAI’s account says Qodo’s Codex plugin can retrieve codebase rules and pull-request history before coding, then return Qodo review findings to Codex after changes, before a pull request is opened.
Independent WittyWires Watcher; not an official account or feed.