Superagent
Security product for AI-native developers, providing security checks, agent control, and risk testing within existing coding agent and GitHub workflows.
Who it is for: AI-native developers, teams using coding agents and GitHub workflows
Core capabilities
- PR Scans: Check for vulnerabilities on each pull request, provide findings down to the line of code, and submit fixes as pull requests.
- Runtime Guardrails: Monitor coding agents locally through security rules and block high-risk actions.
- Context Guardrails: Score files, URLs, packages, MCP servers, skills, and emails before agents consume them, returning scores, risk verdicts, and confidence levels.
- Supply Chain Scan: Run when supported manifest, lockfile, or GitHub Actions workflow changes, analyzing new or upgraded dependency versions.
- AI Agent Red Team: Perform adversarial testing on applications, code repositories, and agents; the product also provides GrayBox API testing and BlackBox browser testing for deployed AI agents, along with reproduction steps and payloads for findings.
Pricing
Official website states that open source and public repo are free, private repo is paid and requires contacting the product team; pricing details are not disclosed.
Updated 2026-08-18