Superagent connects the software you protect to repeatable security operations. It runs built-in security workers, routes findings into your development workflow, and retains structured results as context for later work. People approve consequential changes.

How the factory works

flowchart LR
  Software["Your software"] --> Workers["Security workers"]
  Workers --> Findings["Findings"]
  Findings --> Workflows["Your development workflow"]
  Workflows --> Results["Reviewed results"]
  Results --> Context["Retained context"]
  Context --> Workers
  Workflows --> Approval["Human decision"]

Each security factory combines software targets with security workers. Schedules, integrations, and automation determine when work runs and where results go.

Your software

Software	How Superagent protects it
Repositories	Reviews pull requests and runs repository Red Team assessments
Applications	Tests deployed Web surfaces through browser-based Red Team assessments
Agents	Tests deployed agents and applies Runtime Guardrails where they run
Infrastructure	Coming soon: developer machines, agent hosts, CI runners, and cloud workloads

Your software describes current coverage and planned infrastructure support.

Security workers

Security worker	Use it to
PR Security	Review code, workflow, dependency, and contributor risk before merge
Secure Dependency Updates	Generate dependency update pull requests that pass package policy
Red Team	Find and reproduce attack paths across the software you ship
Agent Guardrails	Check what agents consume and monitor or block risky actions

The Security workers overview describes their coverage.

Documentation map

• Start here
• Your software
• Use cases