AgentSeal

Name: AgentSeal
Author: JoeyBrar

@JoeyBrar

关于 AgentSeal

Every agent action is recorded in a SHA-256 hash chain. With this, you can actually prove to clients that your agent did what it said it did. Record, query, verify, and export agent activity.

#security #audit #compliance #logging #ai-agents #agents

基本信息

分类

其他

传输方式

stdio

发布者

JoeyBrar

提交者

Joey Brar

配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "agentseal": {
      "command": "npx",
      "args": [
        "-y",
        "agentseal-mcp"
      ],
      "env": {
        "AGENTSEAL_API_KEY": "as_sk_your_key_here"
      }
    }
  }
}

工具

未检测到工具

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is AgentSeal?

AgentSeal is an MCP server that provides an immutable audit trail for AI agents, using SHA-256 hash chaining to ensure data integrity. It is designed for multi-agent tracking, on-demand verification, and export, built to meet EU AI Act compliance requirements.

How to use AgentSeal?

Set up in one command: npx agentseal-mcp with your API key. Use the server’s three tools — record_action, query_actions, and verify_chain — to record, look up, and verify agent actions. A Python SDK (pip install agentseal-sdk) is also available for direct integration.

Key features of AgentSeal

One-command setup (npx agentseal-mcp) with API key
On-demand chain integrity verification
Multi-agent tracking under one organization
Export full trail as JSON or CSV
Python SDK for direct integration (agentseal-sdk)
SHA-256 hash chain — tamper-evident records
EU AI Act ready for high-risk AI compliance

Use cases of AgentSeal

Record and verify actions of a single AI agent
Monitor multiple agents across an organization
Provide immutable logs for regulatory audits
Export agent activity trails for analysis or reporting
Ensure data integrity in high-risk AI deployments

FAQ from AgentSeal

What makes AgentSeal different from a regular database?

AgentSeal uses a SHA-256 hash chain where each entry’s hash includes the previous one, so any modification breaks the chain. This provides tamper-evident, verifiable integrity that ordinary databases do not guarantee.

What runtime or dependencies does AgentSeal require?

No additional runtime—just run the one-command npx agentseal-mcp with your API key. A Python SDK is available for direct integration if preferred.

Where does the audit data live?

Data is stored and managed via the AgentSeal service; you sign up at agentseal.io for an API key. The server records and retrieves action data through that service.