The emergence of agentic coding tools is transforming how software development teams create code. A recent study analyzing over 25,000 agent-generated pull requests across thousands of GitHub repositories reveals revealing patterns about how this technology is being adopted. Far from being a massive and immediate adoption, the data shows that most projects generate only one or two agentic contributions per quarter, suggesting that integrating artificial intelligence agents into the development lifecycle is a gradual, nuanced process.
For software development companies, this information has direct implications. It is not enough to have an advanced coding agent; the key lies in designing human oversight processes that maximize its potential without compromising quality or security. The study identifies that the predominant model is individual oversight, where a single developer reviews and modifies the agent’s contributions. Multi-person collaboration patterns are still rare, indicating that organizations are exploring governance models adapted to this new dynamic.
One of the most interesting findings is that small projects (with 1 to 5 contributors) show higher relative participation in agentic activity than medium or large ones. This makes sense: small teams can integrate new tools more quickly and establish agile workflows without the bureaucracy of larger projects. However, productivity measured in PRs per participant varies enormously: only a handful of projects exceed the industry-cited threshold of 36 PRs per participant over three months. Most remain below, suggesting that maturity in agent use is still low.
From a business perspective, these findings reinforce the need for a structured approach to adopting AI in software development. At Q2BSTUDIO, as a company specializing in technology and custom software development, we understand that each organization has a unique pace and set of needs. That is why we offer services ranging from custom software development to the integration of intelligent agents into existing processes. It is not just about implementing a tool, but about redesigning the workflow so that humans and agents collaborate efficiently and securely.
Cybersecurity is another critical factor. When an agent generates code that is later integrated into a repository, it is essential to ensure it does not introduce vulnerabilities. The study notes that human oversight remains the main control mechanism, but as the volume of agentic contributions grows, companies will need automated security analysis tools. At Q2BSTUDIO we offer cybersecurity and pentesting services to validate that AI-generated code meets the highest standards.
Another relevant aspect is cloud infrastructure. Coding agents often run in cloud environments, whether AWS or Azure, and require efficient resource management to scale without skyrocketing costs. The combination of AI agents with well-configured cloud services can significantly accelerate development cycles. Q2BSTUDIO has expertise in cloud services on Azure and AWS, helping companies deploy and optimize their agentic development environments.
Furthermore, agent-generated code produces a massive amount of data on development patterns, frequent errors, and areas for improvement. This data can be leveraged through Business Intelligence tools, such as Power BI, to obtain productivity, quality, and collaboration metrics. A dashboard that visualizes agentic contributions, acceptance rates, and review times enables technical leaders to make informed decisions. At Q2BSTUDIO we implement BI and Power BI solutions that integrate with development platforms to provide a complete view of the software lifecycle.
Process automation is another fundamental pillar. Coding agents not only generate code but can also handle repetitive tasks like unit testing, static analysis, or documentation generation. This frees developers to focus on higher-value tasks. However, automation must be carefully designed to avoid bottlenecks. Q2BSTUDIO offers software process automation services that allow gradual and controlled integration of agents.
In the field of artificial intelligence, the trend is clear: coding agents will evolve rapidly, but their success will depend on organizations’ ability to adapt their governance processes. The analyzed study shows that intensive adoption is still concentrated in a few projects, but that does not mean small companies cannot benefit. On the contrary, small teams have the agility to experiment and find the oversight model that works best. The key is not to underestimate the importance of the human factor: oversight, code review, and usage policy definition are as important as the agent’s technical capability.
Q2BSTUDIO, as a software development and technology company, accompanies its clients on this transformation journey. From initial consulting to implementing artificial intelligence solutions, through cloud integration and cybersecurity, we offer a comprehensive approach. Early adoption of agentic tools is not an isolated experiment; it is a trend that will define the future of software development. Companies that start designing their human-agent collaboration processes now will be better prepared to scale when the technology matures.
In conclusion, the study on 25,264 agentic PRs on GitHub offers an early but revealing snapshot. Adoption is uneven, productivity varies, and individual oversight dominates. However, these data are just the beginning. As agents become more capable and organizations learn to manage them, we will see deeper integration. For custom software companies like Q2BSTUDIO, this is an immense opportunity: we can help our clients not only adopt the technology but do so intelligently, securely, and efficiently.
