The openclaw/openclaw GitHub repository experienced a dramatic rise in pull requests (PRs) from two per week in December to 3,400 per week by February, according to data shared by Greptile. This spike coincided with a sharp drop in PR acceptance rates, falling from about 48% before the surge to under 9.3% afterward. Many of the new PRs were generated by AI coding agents, with one contributor submitting 106 PRs in a single day, averaging just three seconds between submissions.
Rahul Bathija of Greptile, which builds AI agents that review PRs, analyzed this trend while monitoring OpenClaw, the fastest-growing repository in GitHub history. Bathija noted that the flood of low-effort, AI-generated PRs resembled the early 2000s email spam phenomenon. The rapid increase in volume overwhelmed maintainers, leading to a significant decline in the merging of contributions. This pattern highlights the challenges of managing AI-driven contributions in open source projects.
The OpenClaw case offers insight into the evolving dynamics of open source development as AI tools become more prevalent. The surge in automated PRs raises concerns about quality control and the need for new mechanisms such as sender reputation systems to filter contributions effectively. This situation parallels early email spam issues, suggesting that open source communities may need to adopt similar defensive strategies to maintain project integrity.
Greptile’s observations from the OpenClaw repository underline the importance of developing AI-aware moderation tools. The data from early 2026 shows that without such measures, the volume of AI-generated PRs could overwhelm maintainers and degrade project quality. Greptile continues to refine its AI agents to better manage this influx and support sustainable open source collaboration.