NEW YORK, Nov. 18, 2026 (GLOBE NEWSWIRE) -- Qodo, the AI Code Quality and Governance platform, today announced Approval Workflows and Agent Session Review, advancing the autonomous software factory with policy-governed approvals and deeper insight into agent-built software. Approval Workflows let qualifying PRs move forward under engineering standards, while Agent Session Review brings related changes into one guided assessment.
Coding agents can already take a ticket, write code and open pull requests, while AI tools review changes in minutes. But the process still slows at approval, where even low-risk changes wait for human sign-off and engineers must piece together how multiple PRs across repositories fit into a larger feature. As AI-generated code scales, that bottleneck is becoming harder to ignore. Google's 2025 DORA report found that teams with high AI adoption saw 154% larger PRs and 91% longer reviews.
"A software factory is only as good as its quality controls," said Itamar Friedman, CEO and co-founder of Qodo. "Teams have automated how code gets written and reviewed, but approval still runs on whoever has time to open the PR. Qodo now approves the changes that clearly qualify, holds back the ones that don't, and gives engineers the full picture when they need to weigh in. That's how teams get more autonomy without giving up quality."
Auto Approval Workflows (Beta) include:
- Policy-based approval: PRs that meet a team's rules are approved, or labeled "Qodo approved," as soon as Qodo's review completes. Teams start from templates such as Low-risk PRs and scope workflows to a workspace, Git organization or repository.
- Guards: Teams name the repositories and authors Qodo should never approve. Qodo re-checks every new commit and withdraws approval if a PR no longer qualifies.
- PR Triage: One view of every PR's status, with a timeline recording which workflow approved it or which guard held it back.
Agent Session Review (Research Preview) runs from Qodo's Agentic Toolbox, gathering every PR that a coding agent created in a session, across one or more repositories, into a single assessment. It shows where the agent departed from the plan, lays out a plan with merge order and what needs close attention, tying each risk to evidence in the codebase. Engineers check the agent's output as a whole feature before it ships.
Together, these capabilities give teams the controls a high-quality software factory depends on: rules that decide what moves forward automatically, guards that keep critical code with people, and a record of every approval decision.
Auto Approval Workflows are available today in Beta. Agent Session Review is available as a Research Preview. Learn more and get started for free at qodo.ai.
About Qodo
Qodo is the AI Code Quality and Governance platform that enables enterprises to maintain code quality at AI velocity. With self-learning rules, high-precision code review across the SDLC, and centralized governance, Qodo gives engineering organizations the system to govern quality, set standards, and maintain oversight at AI scale and speed. Founded in 2022, Qodo has raised $120 million in funding.
Media Contact
Ashley Hayford
Scratch Marketing + Media for Qodo
qodo@scratchmm.com