September 2025 Summaries

AI-powered coding alone is insufficient for achieving true software delivery velocity, which requires end-to-end automation and intelligent governance across the entire lifecycle. The widespread adoption of AI coding assistants has transformed software engineering, with many organizations shipping code faster; however, this acceleration has exposed a critical bottleneck where downstream processes like testing, security, and compliance remain largely manual, leading to the AI Velocity Paradox. Harness addresses this by providing a unified platform that extends AI beyond code creation, automating building, testing, securing, deploying, and optimizing software. This approach mitigates security and quality risks and reduces cloud costs, while enabling better governance and incident response. By automating the entire software delivery lifecycle, Harness helps organizations become Velocity Leaders, combining speed with resilience to turn AI-driven development into a sustainable competitive advantage, and transforming AI from a productivity boost into a true competitive advantage in the process.

Amazon Kiro and Harness's MCP server integration revolutionizes CI/CD pipeline management by allowing developers to address pipeline issues directly from their IDE using natural language, thereby reducing manual effort and speeding up software delivery. Amazon Kiro, an agentic IDE, goes beyond traditional AI coding assistants by incorporating a reasoning loop that plans, reasons, acts, and evaluates to handle complex tasks, while Harness focuses on post-code generation tasks such as testing, security, deployment, and optimization. The integration enables seamless interaction with Harness's suite of DevOps tools, allowing developers to troubleshoot and fix pipeline failures without leaving their coding environment. This partnership not only streamlines troubleshooting but also enhances pipeline creation, issue detection, and security compliance, ultimately bridging the gap between coding and deployment operations. As the software development landscape evolves, such intelligent integrations are becoming crucial for maintaining development velocity and DevOps excellence.

Harness has acquired Qwiet AI, formerly known as ShiftLeft, to enhance application security in the AI-driven development landscape by integrating agentic AI-powered vulnerability detection and reachability analysis into its platform. This acquisition underscores Harness's commitment to embedding security into the software development process, addressing the challenges posed by rapid AI-fueled coding practices that often introduce vulnerabilities. By incorporating Qwiet's Code Property Graph and its focus on reachability-aware analysis, Harness aims to reduce noise and prioritize real risks, allowing development teams to focus on critical vulnerabilities with AI-guided remediation. This strategic move builds on Harness’s previous merger with Traceable, further consolidating its position in the DevSecOps space by providing a comprehensive, AI-native solution that integrates build, test, deploy, and secure delivery. With these enhancements, Harness seeks to ensure that security is an intrinsic part of the developer workflow, enabling teams to maintain speed without compromising on security in the rapidly evolving AI era.

Harness GitOps extends Argo CD by incorporating the Harness GitOps Agent and integrating with Harness SaaS to enhance governance, audit, and role-based access control (RBAC) while maintaining the core reconciliation loop of Argo CD. This integration allows teams to manage multi-cluster environments with unified dashboards, promotion pipelines, and rollback capabilities, with Git serving as the single source of truth. The architecture involves a control cluster running Argo CD components and the GitOps Agent, which connects to Harness SaaS for visibility and governance, without compromising data security, as sensitive information remains within the Kubernetes namespace. Harness SaaS aggregates data from multiple regions, offering features such as unified dashboards, environment filtering, and multi-cluster promotion pipelines with automated verification and rollback options. This setup enables organizations to scale their GitOps practices while ensuring robust governance and observability across diverse environments. Dewan Ahmed, a Principal Developer Advocate at Harness, brings extensive experience in DevOps and infrastructure, contributing to the tech community through advocacy and coaching.

Harness MCP Server integration with Cursor revolutionizes chaos engineering by making it an accessible and conversational workflow for developers directly within their AI-powered IDE. This integration allows DevOps, QA, and SRE teams to conduct chaos experiments using natural language prompts without needing extensive vendor-specific expertise, thus democratizing resilience testing across various functions. By combining AI-first approaches with comprehensive resilience tools, it simplifies the traditionally complex and steep learning curve associated with chaos engineering, enabling proactive resilience testing as a seamless part of the development process. The integration supports the entire chaos engineering lifecycle, from discovering and executing experiments to analyzing results, thereby fostering a resilience testing culture that is crucial for modern distributed systems. With this approach, organizations can conduct chaos experiments effectively, ensuring robust application development without the burden of deep technical knowledge or context switching.

Harness has been recognized as a Leader in the 2025 Gartner Magic Quadrant for DevOps Platforms for the second consecutive year, highlighting its comprehensive solutions and capabilities in the software delivery lifecycle. Harness attributes this recognition to its strong product strategy, platform breadth, and profound understanding of the DevOps landscape, which has been achieved through an intentional strategy of internal entrepreneurship. The platform supports organizations globally by streamlining software delivery, reducing complexity, and enhancing developer productivity, offering a modular approach that allows for adoption at a pace suited to the specific needs of each organization. Looking forward, Harness aims to further improve the developer experience, simplify DevOps adoption, and integrate security and reliability into the development lifecycle, while its ecosystem of open-source tools and third-party integrations continues to expand.

Harness AI is transforming the landscape of autonomous software delivery by automating maintenance, troubleshooting, and rollbacks, allowing developers to concentrate on innovation. At the Unscripted 2025 event, the focus was on AI's role in reshaping software development and delivery, particularly in overcoming the bottlenecks of transitioning code from development to production. Harness is expanding AI applications across the entire software delivery lifecycle, including Autonomous Code Maintenance (ACM), which automates tasks like dependency upgrades and bug fixes, and AI Verification and Rollback, which simplifies deployment confidence by autonomously setting up health checks and executing rollbacks. New features like Architect Mode and IDP Knowledge Agent aim to assist developers in creating production-ready pipelines and accessing internal platform data through natural language interactions. Harness's advancements in AI-driven software delivery are poised to enhance developer productivity and streamline complex processes in regulated industries.

Database rollbacks are often neglected in deployment processes, yet testing undo migrations is crucial for safe database changes. Harness Database DevOps, in combination with Liquibase, automates rollback testing, allowing teams to implement changes with increased confidence. Unlike application code, where rollbacks and testing are standard, database rollbacks are more complex and frequently left untested. Undo migrations are scripts that reverse database schema changes, and their testing is essential to prevent failures due to schema drift, dependencies, or overlooked edge cases. Harness Database DevOps applies CI/CD principles to database pipelines, ensuring consistency and pre-deployment validation of rollbacks, which are tested in a controlled environment before hitting production. A typical workflow involves using Liquibase to check history, take database snapshots, apply schema changes, and then roll them back to compare snapshots, ensuring rollback efficacy. This integration with Liquibase not only automates rollback testing but also makes schema changes more reliable and less stressful for developers and DBAs.

Harness MCP Server and Windsurf integration simplifies chaos engineering by using natural language prompts to conduct experiments, making resilience testing more accessible for DevOps, QA, and SRE teams. This approach reduces the learning curve and technical complexity traditionally associated with chaos engineering, enabling teams to execute and analyze experiments without needing extensive vendor-specific knowledge. The integration offers a suite of tools that cover the entire chaos engineering lifecycle, allowing for the discovery, execution, and monitoring of chaos experiments directly within the development environment. This AI-powered solution not only democratizes chaos engineering but also encourages a culture of resilience testing, helping organizations build robust applications capable of withstanding production challenges.

Comet, a new autonomous browser from Perplexity, excels at quickly performing web tasks by relying on the accessibility tree and raw DOM for interacting with web elements, but it faces limitations due to the inherent messiness and diversity of the web. While Comet is effective on well-structured sites that adhere to best practices, it struggles with complex applications that do not follow standard web development rules, particularly when dealing with cross-origin iframes or elements lacking proper labels. Comet's reliance on HTML structures and accessibility features, without utilizing screenshots, presents challenges when interacting with images or icons that aren't properly labeled. In contrast, the Harness AI Test Automation employs a bimodal approach, combining screenshots with DOM and accessibility data to provide broader coverage and more reliable interactions, especially in enterprise applications with older or inconsistent accessibility implementations. Ultimately, Comet is well-suited for environments where its assumptions are met, but a bimodal approach could enhance its effectiveness across the diverse landscape of the web.

Harness Chaos Engineering is a module within the Harness platform designed to enhance system resilience through proactive testing and simulation of real-world failures. It supports various teams, including developers, QA engineers, and Site Reliability Engineers, by uncovering potential weaknesses in infrastructure and offering actionable insights via resilience scores and test coverage metrics. The platform provides a comprehensive set of tools for end-to-end resilience testing, featuring an extensive library of chaos faults, automated resilience probes, and AI-driven recommendations. Key use cases include integration with deployment pipelines, combining with load testing, and conducting disaster recovery exercises, all aimed at turning vulnerabilities into opportunities for improvement. Harness Chaos Engineering offers flexible deployment options, including a free SaaS plan and an on-premise solution, catering to different organizational needs and compliance requirements, thereby enabling businesses to improve their infrastructure's reliability and resilience effectively.

On September 8, 2025, a significant supply chain incident impacted the open-source community when attackers compromised a maintainer account and injected malicious updates into over 18 popular NPM packages, such as chalk and debug, which aimed to steal cryptocurrency wallets. The incident highlights the vulnerability of the software ecosystem to a single point of compromise and underscores the necessity for organizations to bolster their defenses through rigorous open-source dependency management and policy-driven security controls. Harness Supply Chain Security (SCS) offers solutions by providing tools for OSS search, AI-assisted policy enforcement, and continuous remediation tracking, allowing teams to quickly identify vulnerabilities, prevent compromised packages from entering new builds, and ensure complete risk mitigation. These capabilities are designed to enhance the integrity of the software supply chain, enabling teams to swiftly detect, prevent, and remediate threats.

Platform engineering has emerged as a pivotal discipline in modern software delivery, transforming how organizations build, deploy, and scale applications. It addresses the limitations of traditional DevOps by offering structured, developer-friendly infrastructure that enhances innovation, security, and cost-efficiency. Gartner predicts a significant rise in platform engineering teams by 2026, driven by the need to manage complex cloud-native architectures. Unlike the ad-hoc approaches that create silos and hinder innovation, platform engineering treats infrastructure and developer tooling as products, establishing "golden paths"—standardized ways to streamline development processes. Harness, a leading platform in this domain, offers a comprehensive, modular solution integrating various aspects of software delivery, including CI/CD, feature management, and security testing. By automating governance and compliance, it ensures security becomes an invisible yet integral part of development. The platform's intelligent pipeline orchestration and reusable templates facilitate seamless deployment and operational excellence, while DORA metrics and adoption tracking provide actionable insights into delivery performance. Despite challenges like achieving developer adoption and balancing flexibility with governance, organizations that embrace platform engineering can enhance their developer experience and maintain a competitive edge. Martin Reynolds, a seasoned expert in software development and DevOps, advocates for this transformative approach, emphasizing its role in fostering collaboration, automation, and continuous improvement in software delivery.