Superagent Technologies
Superagent builds security for AI-native developers: software that helps teams trust what they merge, control what their agents do, and demonstrate that their software is safe. Its autonomous security factory protects applications, agents, repositories, and infrastructure from development through production.
At a Glance
- AI-native software developers and engineering teams
- Open-source maintainers
- Companies shipping AI agents and agentic applications
- Security and platform engineering teams
- +2 more
AI Tools by Superagent Technologies
(3)grok-cli
Open Source Grok Terminal Agent
Superagent SDK
AI Runtime Safety SDK
Vibekit
AI Agent Security Sandbox
Discussions
No discussions yet
Be the first to start a discussion about Superagent Technologies
Latest News
Products & Services
Runtime and software-security platform that analyzes requests and responses, detects prompt injections and jailbreaks, filters secrets and PII, safeguards tool/code execution, and provides traces, audit trails, and policy controls.
GitHub-native pull-request scanning that blocks risky code and untrusted contributors before merge, with findings delivered as check runs and PR comments.
Scheduled dependency remediation using pinned Dependabot tooling in an isolated sandbox; Superagent evaluates package changes and opens only policy-cleared pull requests.
Automated adversarial testing for repositories, applications, agents, and infrastructure that finds attacker exploit paths and supplies reproducible findings and fixes.
Market Position
Superagent positions itself as an autonomous, AI-native security factory rather than a point scanner: it spans PR security, dependency remediation, red teaming, context and runtime guardrails, and agent workflows while preserving a customer's existing integrations. VibeKit operates in the coding-agent sandbox market alongside Modal, E2B, Daytona, Cloudflare, Vercel, Beam, and Blaxel; Superagent differentiates with security controls, redaction, policy, and observability layered across the agent lifecycle.
Leadership
Founders
Alan Zabihi
CEO and co-founder. Together with Ismail Pelaseyed, he has built several companies over more than 15 years; the pair have worked with Google, SKF, Klarna, and Trustly, published CVEs in widely used open-source projects, and built an early open-source AI-agent framework adopted by enterprises.
Ismail Pelaseyed
CTO and co-founder. He has worked with Alan Zabihi for more than 15 years across several companies, including work with Google, SKF, Klarna, and Trustly; the co-founders have published CVEs and maintained open-source agent and security tooling.
Executive Team
Alan Zabihi
Chief Executive Officer and Co-founder
Has built several companies with Ismail Pelaseyed over 15-plus years; worked with Google, SKF, Klarna, and Trustly; published CVEs and helped build an early open-source AI-agent framework.
Ismail Pelaseyed
Chief Technology Officer and Co-founder
Long-time co-founder with Alan Zabihi; worked with Google, SKF, Klarna, and Trustly and has contributed to open-source AI-agent and security tooling, including published CVEs.
Founding Story
Superagent was started because AI-native development was moving faster than manual security review: agents write and merge code, consume untrusted context, and operate with access that people cannot continuously watch. The founders' initial vision was to put security directly where developers already work, combining reasoning-driven defenses with open-source tools, agent sandboxing, and deterministic guardrails.
Business Model
Revenue Model
Usage-based SaaS for private repositories, sold in monthly or annual credit plans; every plan includes the feature suite, API and MCP access, and unlimited seats. Public/open-source repositories are free.
Pricing Tiers
Unlimited usage on public repositories across the Superagent suite.
100 private credits per month.
500 credits per month.
2,000 credits per month.
6,000 credits per month.
Target Markets
- AI-native software developers and engineering teams
- Open-source maintainers
- Companies shipping AI agents and agentic applications
- Security and platform engineering teams
- Organizations with GitHub-based development and CI/CD
- Teams running coding agents against private or production codebases
- Always-on repository penetration testing with Cursor
- Validate dependency updates end to end with Devin
- Security triage for Supabase applications
- Secure pull requests and untrusted contributors before merge
- Protect coding agents operating on sensitive codebases
- Score untrusted packages, pages, MCP servers, and messages before agents consume them
- Ubicloud
- Evry Health
- dotenvx
- Firecrawl