Qodo Automates Code Approvals With Guardrails, Not Free Rein
Qodo's beta Auto Approval Workflows, announced November 18, let its bot approve routine PRs as one vote while required reviewers stay in place; a research preview consolidates multi-repo agent sessions.
By Olivia Hart
4 min read

What's News
- Qodo announced Auto Approval Workflows (beta) and Agent Session Review (research preview) on November 18.
- Qodo's bot approval counts as one approval; required reviewers and code-owner rules still apply.
- Each new commit triggers reassessment, and approval is withdrawn if the PR no longer qualifies.
- Agent Session Review consolidates pull requests from multi-repository coding sessions into one assessment.
Qodo now lets its bot approve routine pull requests — but that approval counts as exactly one vote, and every required human reviewer still has to sign off. The company announced the move on November 18, shipping Auto Approval Workflows in beta alongside Agent Session Review, a research preview that consolidates multi-repository AI coding sessions into a single assessment.
The release targets two bottlenecks in AI-assisted development. Routine pull requests often sit in a queue waiting for a human approval even after automated review has finished. Larger agent-built features create the opposite problem: several individually reasonable changes scattered across repositories can be hard to read as one coherent implementation. Qodo's answer pairs configurable approval policies with a consolidated view of an agent's complete session.
What does the approval workflow actually change?
A pull request proposes a change to a codebase. Reviewing it identifies problems; approving it records the decision that the change can move forward. Automating the first step, Qodo notes, does not automatically clear the queue at the second.
According to Qodo's announcement, Auto Approval Workflows lets teams set rules for qualifying changes. When Qodo finishes reviewing a PR, a matching workflow can approve it or apply a "Qodo approved" label. Teams can start with templates for low-risk work and apply workflows at the workspace, Git organization or individual repository level.
The label-versus-approval distinction matters. A label signals quality without authorizing a merge. In its product announcement, Qodo says its bot's approval counts as one approval, while existing required reviewers and code-owner requirements still apply. The feature is not unrestricted automatic deployment, and Qodo does not claim it is.
What changes is where the repeatable decision gets made. Instead of an engineer rediscovering the same acceptance criteria for every routine update, the team encodes those criteria in a workflow. The value hinges on whether the policy captures what actually makes a change acceptable — not on how many approvals the system issues.
How does Qodo keep humans on critical code?
Guards let teams designate repositories and authors that should never receive automatic approval. Qodo says each new commit triggers reassessment, and approval is withdrawn if the updated pull request no longer qualifies. A PR Triage view records the approval status and the workflow or guard behind each decision.
That reassessment is not a formality. Approval is a judgment about one version of the code; a small fix can grow into a larger change when commits arrive. An approval that survives those changes unchanged would say less about the final implementation than it appears to.
The controls also make automation's boundary visible. A team can delegate a narrow class of routine decisions while retaining direct review everywhere else — a narrower proposition than asking engineers to trust all AI-generated code, and one that leaves policy ownership with the deploying organization.
Can a whole AI coding session be reviewed at once?
Agent Session Review tackles work that no single pull request can capture. Running through Qodo's Agentic Toolbox, it gathers the PRs produced during a coding session — including work spanning multiple repositories — into one assessment. Qodo says the result connects implementation decisions, deviations from the original plan, merge order and risks to evidence in the codebase.
For an engineer, the useful unit of review may be a complete feature's behavior rather than repository boundaries. A backend change and a client update can each look reasonable in isolation while resting on conflicting assumptions about how they interact. A session-level assessment is built to make those relationships inspectable.
The research-preview label carries weight. Qodo is describing a new way to organize evidence for review, not demonstrating that every unintended behavior gets caught. Engineers still decide whether the implementation matches the intended requirements.
Where does this leave software autonomy?
The two features point to a division of labor: policies handle changes whose acceptance conditions can be stated clearly, while people examine decisions and interactions that need broader judgment. The announcement provides no independent evidence of faster delivery or fewer defects, and those outcomes should not be assumed from automated approval alone.
The larger shift for development teams is from measuring how fast an agent writes code to understanding how that code earns permission to proceed. Qodo's beta workflows and session-review preview move that permission process closer to the center of the AI development toolchain — with the burden of writing good policy now sitting squarely with the teams that deploy it.
Source: unite.ai
More from Olivia Hart
Show full bio
Staff writer covering people, launches and day-to-day practice in entrepreneurship, startups and venture funding at Business Bearings.
2,378 articles