New Delhi, August 31(APAC Media): Harness has launched a new code repository and AI-powered code review tools as software teams increasingly use AI agents to write and submit code.
The company introduced Agent-Ready Harness Code Repository and AI Code Review, targeting a problem emerging with agent-based software development: the growing volume of code and pull requests that developers need to manage and review.
AI Agents Are Changing Code Workflows
AI coding agents can generate and submit code at a much faster pace than traditional developer workflows. This can increase the number of commits, branches and pull requests that development teams need to process.
Harness said its new Code Repository is designed to handle higher repository activity, including large numbers of simultaneous commits and pull requests. It also adds permissions that allow organisations to define what individual AI agents can access, merge or deploy.
The repository supports role-based access control and Open Policy Agent policies, giving teams more control over non-human identities working on software projects.
AI Code Review Focuses On Higher-Risk Changes
Alongside the repository, Harness has introduced AI Code Review to reduce the amount of manual review required for AI-generated code.
The system groups changes based on risk rather than simply listing modified files. It can also provide suggested reviewers and labels and allow teams to set review checks for pull requests.
Harness said the review system can use information from its Software Delivery Knowledge Graph, including previous software incidents, to provide additional context when assessing code changes.
The company is also positioning the tools as part of its wider software delivery platform. Harness said its engineering teams have used the system internally and recorded more than 10,000 hours of manual review time saved per month.
The new repository is available with a free tier offering 50GB of storage. Harness also supports migration from GitHub, GitLab, Bitbucket and Azure DevOps, including repositories, pull requests and selected development workflows.
The launch comes as software companies increasingly add AI agents to coding and DevOps processes, putting greater focus on how organisations review, govern and deploy machine-generated code.
