Superagent

Superagent is a security platform for AI-native developers that works in coding-agent and GitHub workflows. It helps teams assess what they merge, red-team applications and agents, and apply runtime guardrails to control agent actions and consumption. The current product includes pull-request checks, security testing, agent guardrails, safety documentation, and developer-facing APIs and integrations. Public open-source repositories have unlimited free usage; private repositories have a monthly free credit allowance and published paid plans. This is an AI security product, not an agent-building framework.

✨ Key Features

• Pull-request security checks — Review AI-related code changes before merge.
• AI application red teaming — Test applications for safety and security weaknesses.
• Agent red teaming — Probe agent behavior and failure modes.
• Runtime guardrails — Control what agents can do while operating.
• Consumption controls — Monitor and constrain agent resource use.
• GitHub workflow integration — Work where developers already review code.
• Coding-agent workflow — Bring security checks into existing coding-agent usage.
• Safety reporting — Produce evidence that an AI product is safer to use.
• Developer APIs — Integrate security capabilities into custom workflows.
• Open-source support — Offer the platform free for open-source projects under current terms.

💳 Pricing

• Public open-source repositories — Unlimited free usage across the suite.
• Private Free — US$0, 100 credits/month.
• Starter — US$499/month, 500 credits/month.
• Growth — US$1,499/month, 2,000 credits/month.
• Scale — From US$3,999/month, 6,000 credits/month.
• Annual billing — Paid annual plans are prepaid and add 20% credits, refreshed monthly; this is a credit bonus, not a stated 20% price discount.
• Plan access — All plans include the products, API/MCP and unlimited seats. PR scans cost one credit, red-team reports 500 and context checks one per item.

⚖️ Alternatives Compared

🎯 Use Cases

• PR review: Check AI-related changes before they reach production.
• Agent security testing: Red-team autonomous workflows.
• Runtime governance: Limit risky agent actions and resource consumption.
• GitHub security: Add AI security checks to existing repositories.
• AI product launch: Gather evidence before exposing an agent to users.
• Open-source maintenance: Add security review to open-source AI projects.
• Developer platform integration: Embed checks into internal tooling.
• Compliance evidence: Document testing and control coverage for stakeholders.