February 2026 Summaries

In light of recent developments in AI, the software development landscape is undergoing a significant transformation, with the traditional bottleneck of coding being replaced by the challenge of managing parallel agents and system architecture. The article discusses the launch of background-agents.com, a platform designed to explore and provide clarity on the concept of 'background agents' and their role in automating new bottlenecks within the software development lifecycle. As AI continues to evolve, the emphasis is shifting towards using cloud-based, autonomous agents to handle tasks like code review and vulnerability remediation, prompting a need for new terminology and understanding in the industry. The website aims to serve as an interactive guide to these emerging trends, while highlighting the importance of adapting to a self-driving codebase to keep pace with exponential technological growth.

Ona Automations offers a solution for planning large-scale git migrations by quickly analyzing hundreds of repositories and generating a migration-ready inventory in minutes. This automation tool tackles the complexities of repository interdependencies, runtime stacks, build tools, CI/CD configurations, and the individuals responsible for maintaining them. It provides a structured report that includes a summary of each repository's function, its technological stack, tools, dependencies, and key personnel involved in its maintenance. By visualizing the dependency graph, teams can efficiently sequence migrations, prioritizing repositories with no inbound dependencies and coordinating those with multiple dependencies. Validated against real-world organizations, the automation significantly reduces the time and effort required for migration planning, transforming a weeks-long manual process into a streamlined operation. Available in the Ona Template Library, this tool can be customized and scaled to meet specific migration needs, helping teams build comprehensive migration plans swiftly.

Ona Automations offers a sophisticated AI-driven solution for automating code reviews, addressing the bottlenecks in the software development lifecycle caused by traditional human-dependent reviews. Unlike other AI tools that only analyze code diffs, Ona's system runs the code in a full development environment, executes tests, and verifies that changes meet project requirements by integrating with tools like Linear, Jira, and Confluence. This approach allows the agent to provide in-depth, actionable feedback, significantly reducing the time spent on pull requests by 50% and achieving an 88% acceptance rate on its suggestions. By operating on the user's infrastructure and ensuring code does not leave the network, Ona's automation enhances security and offers extensive customization options, allowing teams to tailor the system to their specific coding standards and project needs. This results in a more efficient shift of human focus from basic issue detection to more complex architectural and design decision-making, thus accelerating engineering processes and improving overall project quality.

Ona Automations introduces proactive background agents that automate software engineering tasks by combining AI prompts with deterministic shell scripts in trigger-based, closed-loop workflows. These agents, now generally available, are designed to handle a variety of tasks such as autonomous backlog picking, Sentry issue triage, codebase cleanup, CVE remediation, and large-scale migrations. By running in secure, fully configured environments, Ona can execute tasks across multiple repositories in parallel, providing a seamless integration with existing tools like Linear, Sentry, GitHub, and GitLab. The system allows engineers to focus on more complex problems by autonomously managing routine tasks, delivering merge-ready pull requests, and offering comprehensive reports for compliance and security audits. Ona Automations can be hosted in the cloud or on enterprise networks, maintaining the same connectivity and context as human engineers, and enabling engineers to take over the automated environment as needed.

Ona Automations offers a solution for managing engineering backlogs by employing an AI-driven process that autonomously handles straightforward tasks such as backend refactors, small bug fixes, and minor feature requests. The system operates on a schedule, scanning project management tools like Linear or Jira, and assesses which tasks an AI engineer can complete. It then selects suitable tickets, executes the tasks in isolated environments, and submits a pull request for human review, ensuring no direct impact on production environments. This method focuses on tasks that are well-defined, backend-oriented, and do not require manual quality assurance, allowing teams to clear backlogs efficiently while maintaining human oversight for quality assurance. Ona Automations is designed to handle one ticket per cycle initially, gradually expanding its criteria as trust in its output grows, ultimately transforming backlogs from neglected collections of tasks into active queues processed by AI engineers.

Ona's engineering team has streamlined the feature delivery process by leveraging automation to minimize manual work and improve efficiency. Traditionally, feature delivery involves a lengthy cycle with multiple handoffs, leading to delays not due to coding complexity but due to procedural bottlenecks. Ona addresses this by utilizing an AI-driven interactive agent that transforms feature specifications into structured plans, auto-generates projects with tickets and dependencies, and executes tasks sequentially. This process reduces context-switching and manual ticket creation, allowing engineers to focus on judgment-based decisions. Ona Automations further enhance efficiency by handling pull request formatting, conducting automated code reviews, and enforcing quality checks, which ensures consistent standards and reduces the need for human intervention until critical decision points. This approach has significantly increased productivity, as evidenced by Ona's engineering team achieving 10.8 pull requests per engineer per week, and paves the way for future autonomous backlog execution.

Ona's engineering team has achieved remarkable productivity, reaching the 99th percentile by leveraging extensive automation, which has allowed them to maintain a high output of pull requests without increasing team size. The key strategy involves automating tasks that do not require human judgment, thus freeing up engineers to focus on tasks that do. Using Ona Automations, the team has systematized their workflow, including planning, execution, and maintaining code quality and hygiene, with automations such as stale pull request cleanup, feature flag removal, and documentation updates. This systematic approach has significantly reduced the time from planning to feature completion, allowing engineers to wake up to new features ready for review. Ona has made these automation tools and methods available to other teams through their Ona Template Library, offering a pathway for other organizations to replicate their success and improve efficiency by automating routine engineering tasks.

In a large monorepo where AI agents collaborate with humans, unused code can degrade AI performance by wasting tokens and creating a noisier context window. To address this, the engineering team automated the cleanup process using Ona Automation and the Knip tool, which scans for unused dependencies, exports, and files. Instead of risky and hard-to-review large pull requests, the automation runs Knip daily, selects one cleanup task, validates it, and creates a small pull request with auto-merge enabled. This approach allows engineers to start their day by approving minor changes, resulting in a continuously cleaner codebase without the need for specific time allocations or cleanup sprints. This method not only enhances agent performance by maintaining a tighter context window but also serves as a model for other automatable tasks that typically lack prioritization, such as JavaScript to TypeScript conversions and dependency updates.

Benjamin Stark discusses the benefits of using Ona Automations to maintain healthy codebases by automating routine tasks such as remediating Common Vulnerabilities and Exposures (CVEs), closing stale pull requests, removing unused feature flags, and updating documentation. These automations operate in secure, isolated environments and function without human involvement, addressing issues that typically accumulate due to competing priorities like feature development and deadlines. By running scheduled or event-triggered tasks, Ona Automations help prevent codebase degradation, reducing the risk of merge conflicts, complexity, and outdated information that can slow down engineering teams. This approach allows developers to focus on work that requires human judgment, while repetitive maintenance tasks are handled automatically, ensuring codebases remain clean and efficient.

The transition from traditional development environments to cloud-based systems is becoming essential due to the rise of background agents that require isolated, fully provisioned environments to function effectively. Companies like Stripe and Ramp have already adopted standardized cloud development environments, enabling them to integrate agents seamlessly into their workflows, resulting in significant productivity gains. These cloud environments address long-standing issues such as environment drift, lengthy onboarding, and reproducibility challenges, which local setups struggle to resolve. As AI-driven agents become more prevalent, the need for cloud-based environments that offer isolation, declarative setup, automated lifecycle management, and secure access to internal resources has become crucial. The shift to cloud environments not only supports the scaling of autonomous agents but also enhances human productivity, security, and collaboration across organizations. The investment in such infrastructure is proving to be a major advantage for companies prepared for this shift, as it allows for greater efficiency and innovation in software development processes.

Benjamin Stark describes the development and implementation of an Ona Automation that addresses the common issue of outdated documentation in engineering organizations by scanning the codebase daily, detecting changes affecting user-facing behavior, and automatically drafting a pull request with updated documentation. This automation contrasts with existing tools like Mintlify, which only flag outdated documentation, by providing a proactive solution that creates draft updates in the company's established tone and style, thus reducing the time engineers spend on manual documentation updates. The system operates in a secure environment, utilizes the same Dev Container setup as the engineering team, and has been fine-tuned to batch changes daily while ignoring irrelevant internal modifications. As a result, the organization experiences improved documentation quality and consistency, reducing the burden on engineers and enhancing customer trust. The automation runs seamlessly in the background, offering a scalable pattern for teams where documentation and code repositories tend to drift apart over time.

Ona has developed an automation solution to address the issue of noisy Sentry alerts, which create alert fatigue and financial costs due to increased event volume. This automation, known as Ona Automation, automatically triages Sentry issues, makes small code fixes, and prepares pull requests (PRs) for review, allowing engineers to focus on product backlog rather than managing monitoring noise. By running every morning, the automation helps maintain a clean Sentry dashboard, reducing direct costs and improving operational efficiency by keeping alerts actionable. The system is designed to be incremental, addressing issues daily to prevent accumulation, and is built as an Ona skill, allowing for manual or scheduled use within secure environments. The approach not only saves costs but also ensures sustainable on-call rotations and cleaner alerts, making it beneficial for both small teams and enterprise organizations facing similar challenges with alert noise.

Ona for Open Source is a program aimed at supporting open-source maintainers and contributors by providing up to $200 per month in free AI credits to tackle challenges such as PR backlogs, contributor onboarding, and context reconstruction during code reviews. Originating from Gitpod, which began as an open-source project, Ona integrates cloud development environments with AI agents that autonomously assist with tasks like fighting AI slop, automating dependency updates, and managing security patches. By addressing persistent problems such as poorly contextualized PRs and "works on my machine" issues, Ona for Open Source seeks to streamline the maintenance of open-source projects, allowing maintainers to focus more on value-driven activities rather than routine tasks.