Get started
Upload an SBOM, read the findings and record your first VEX statement in 15 minutes: Your first release. Then add a unit, or see how Affectra fits your process.
Affectra takes the SBOMs your CI builds produce and matches their components against vulnerability data, so that every release has a list of findings. You record what each finding means for your product as a VEX statement. A collector on the units in the field reports which packages are installed, so you can see which units are exposed.
Affectra is built to work with AI assistants. Connect Claude, Codex or another MCP client, and it ranks your findings, explains how a vulnerable component gets into a release, finds the units a vulnerability reaches, and proposes VEX statements backed by evidence from your data. A person in your organization decides. Built for AI assistants explains how.
Get started
Upload an SBOM, read the findings and record your first VEX statement in 15 minutes: Your first release. Then add a unit, or see how Affectra fits your process.
Work with an AI assistant
Connect an AI assistant over MCP in a few minutes, signed in as you or with a read-only key. Then let it propose statements and review them. The tools are listed in the MCP server reference.
Guides
Steps for one task. Start with uploading SBOMs from GitHub Actions, or installing the collector on a unit. Then assess your findings.
Reference
The words Affectra uses, and what each one means. Start with the glossary, the VEX statuses or the vulnerability sources.
Concepts
How Affectra works. Start with built for AI assistants, how matching works, or why a finding is or isn’t shown.
Security and compliance
The licences and notices for third-party data that Affectra shows.