superagent
Superagent protects your AI applications against prompt injections, data leaks, and harmful outputs. Embed safety directly into your app and prove compliance to your customers.
The Superagent MCP server integrates AI application safety features into Model Context Protocol-compatible clients like Claude Desktop. It provides runtime protection against prompt injections, automatic redaction of personally identifiable information and secrets, and repository scanning for AI-targeted attacks. Developers can embed these security controls directly into their AI workflows to block malicious instructions, prevent data leaks, and demonstrate compliance. The server connects to the Superagent safety-agent SDK, which works with any language model provider.
Key Features
Use Cases
- 01Protecting AI assistants from prompt injection attacks that attempt to override system instructions
- 02Automatically scrubbing sensitive data like emails, SSNs, and API keys from agent inputs and outputs
- 03Scanning third-party repositories before agents clone or reference them to prevent supply-chain attacks
- 04Ensuring AI applications meet compliance requirements by demonstrating active security controls
- 05Running security audits on agent endpoints to identify vulnerabilities before production deployment
- 06Blocking data exfiltration attempts where attackers try to extract training data or system prompts
Related MCP Servers
View moremarkitdown
Python tool for converting files and office documents to Markdown.
firecrawl
The context API to search, scrape, and interact with the web at scale. 🔥
prompts.chat
f.k.a. Awesome ChatGPT Prompts. Share, discover, and collect prompts from the community. Free and open source — self-host for your organization with complete privacy.
langflow
Langflow is a powerful tool for building and deploying AI-powered agents and workflows.
superagent — FAQ
What is the Superagent MCP server?+
The Superagent MCP server is a safety layer for AI applications that integrates with Claude Desktop and Claude Code via the Model Context Protocol. It provides guard, redact, and scan capabilities to protect against prompt injections, data leaks, and malicious repository content.
How do I install the Superagent MCP server?+
Installation requires signing up at superagent.sh to obtain an API key, then adding the MCP server configuration to your Claude Desktop or Claude Code settings. You can use either the TypeScript (npm) or Python (uv) SDK packages to run the server.
Which AI clients work with this MCP server?+
The Superagent MCP server is designed for Claude Code and Claude Desktop, as documented in the integration options. These are the officially supported MCP-compatible clients.
Do I need an API key to use Superagent?+
Yes, you need to sign up at superagent.sh to obtain a SUPERAGENT_API_KEY. This key must be set as an environment variable before using the guard, redact, or scan features.
Is the Superagent MCP server free?+
The software is open source under the MIT license, but usage requires a Superagent API key obtained from superagent.sh. Pricing details are available on their website, and the scan feature shows per-operation cost information.
Can I run Superagent on my own infrastructure?+
Yes, Superagent offers open-weight models that can run on your own infrastructure with 50-100ms latency. This allows you to keep sensitive data within your environment while still using the guard functionality.
How do I install superagent?+
Open the source repository on GitHub and follow its README. superagent is a mcp server — MCP Agents Market links you directly to the official repo.
Is superagent free?+
superagent is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.