Research checked: November 18, 2026. This is a documentation-based tool profile, not a hands-on benchmark or a new-release announcement.
Hayabusa, maintained by Yamato Security, helps incident responders examine Windows event logs and build investigation timelines. Its Sigma-based detections provide starting points for threat hunting, while CSV, JSON and JSONL output can support further analysis. The project documents integrations with tools including Timesketch and Velociraptor.
What it analyzes, and where it runs
The evidence is Windows event logs; the analyst’s computer does not have to run Windows. The official downloads documentation lists Windows, macOS and Linux binaries for several processor architectures. That makes it useful for teams collecting Windows logs centrally or examining exported evidence on a separate workstation. It should not be mistaken for a general Linux audit-log detector.
Licensing has two parts
The project’s license documentation identifies GNU AGPLv3 for the Hayabusa program and Detection Rule License 1.1 for the detection rules. Keep those components separate when evaluating redistribution or integrating the software into a service. An open repository does not mean that every bundled component uses the same terms.
A useful first evaluation
Start with preserved log copies from a controlled test computer. Record where the logs came from, the collection time and the timezone used in the output. Generate a timeline, locate several events whose circumstances you already know, and compare the result with the original records. Then inspect the evidence behind a small number of detections before processing a larger collection.
Keep exported timelines within the investigation’s access controls: log records can contain account names, device information and other sensitive details. Preserve the original evidence separately from working files so another analyst can reproduce the review.
Know the limits
A detection is an investigative lead, not a verdict. Disabled, overwritten or uncollected logs leave gaps that a timeline cannot reconstruct. A quiet result does not establish that a machine was never compromised. Hayabusa is most useful when its findings are checked against endpoint, identity and network evidence, with the scope and limitations of the collection documented.
Related Digital Arsenal research
Velociraptor covers endpoint evidence collection; Sigma covers detection-rule research; Timesketch covers collaborative timeline analysis.
Featured image: conceptual IPCRIME artwork, not a photograph or screenshot of this incident or product.
