AgentLogs makes it possible to study the sequence of actions behind a coding AI agentAn AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.Full definition →'s contribution. The dataset contains 549,239 sessions from GitHub's Copilot cloud-agent platform across 35,810 public repositories. Its logs include prompts, model output, tool calls and usage records, allowing researchers to connect a task with the searches, edits and commands used to address it. Multiple sessions can belong to one task. That structure supports questions about repeated attempts, tool useA model's ability to call external functions — run code, search the web, edit files — instead of only generating text.Full definition → and resource consumption that a final pull request alone cannot answer. Dataset overview and structure.
Terms in this piece · Glossary
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
tool use — A model's ability to call external functions — run code, search the web, edit files — instead of only generating text.
token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
articleNot All Agents Are Equal: Code Quality and Post-Merge Maintenance Across Five Autonomous Coding Agents in the WildObada Kraishan
articleLabAgent: Customize Any Research Hubs for Scientific Discoveries Using AI AgentsLei Liu, Yikun Zhang, Jialin Chen, Wanjia Zhao, Rex Ying, Wengong Jin, Hua Xu, James Zou, Tianyu Liu, Hongyu Zhao
articleXAgent: eXecution-guided Agentic AI for Effective Localization and Resolution of GitHub IssuesHieu Huynh, Patanamon Thongtanunam, Michael Fu, Bach Le, Kla Tantithamthavorn