Do These Violent Delights Have Violent Ends?
Do These Violent Delights Have Violent Ends?
Measuring the Post-Merge Fate of Agentic Code
Paper<br>Read the manuscript -->
GitHub<br>View replication package -->
First longitudinal empirical study of the post-merge lifecycle of agentic and human code in real-world open source projects.
182<br>public GitHub repositories
30K+<br>agentic commits
121M+<br>lifecycle lines tracked
12<br>maintenance operational intents
143K+<br>security & dependency findings
1 year<br>May 2025 to May 2026
Finding Highlights:
46%<br>higher corrective maintenance rates than human contributions
1.51x<br>times more high-severity static security findings than human contributions
Agentic code share<br>increases the corrective maintenance burden within projects
No review rate<br>increases the agentic maintenance burden of a project
Interested to learn more? Check out the interactive visualization to play with the data yourself!
Citation:
@article{postmerge26,<br>title = {Do These Violent Delights Have Violent Ends?<br>Measuring the Post-Merge Fate of Agentic Code},<br>author = {Xia, Chunqiu Steven and Miller, Courtney},<br>year = {2026},<br>journal = {arXiv preprint arXiv:temp},<br>url = {https://arxiv.org/pdf/temp}<br>Copy
Search repo
Language
Repo focus
Min. agentic commits
Reset filters
Repo-Week Timeline
Weekly tracked-line survival rates for agentic and human origin lines.
Commit Contribution Maintenance Class Mix
Origin-to-Terminal Actor Flow
Line-Level Termination Maintenance Class Proportion
Semgrep Findings
OSV-Scanner Findings
Repositories
Filtered repository summaries.
Repo<br>Language<br>Stars<br>Agentic Commits<br>Agentic Share<br>Latest Agentic Lines<br>Survival