Keeper Security Announces "Agent Kit" to Secure AI Development Workflows
Keeper Security APAC has announced "Agent Kit," a solution to secure AI development workflows. It enables AI agents to safely manage privileged credentials, reducing the risk of sensitive information leakage during the development process.
📋 Article Processing Timeline
- 📰 Published: April 30, 2026 at 18:00
- 🔍 Collected: April 30, 2026 at 09:31
- 🤖 AI Analyzed: April 30, 2026 at 09:41 (10 min after Collected)
Keeper Security APAC Inc. (Asia Pacific Headquarters: Tokyo, CEO and Co-founder: Darren Guccione, hereinafter "Keeper"), a cybersecurity provider known for zero-trust and zero-knowledge based identity security and privileged access management (PAM), announced the launch of "Agent Kit," which enables safe automation of tasks in AI development environments.
Agent Kit is a suite of features that directly integrates Keeper Secrets Manager and Commander with leading AI coding agents such as Claude Code, Cursor, Codex, and GitHub Copilot. It allows for the secure automation of security settings and management tasks.
In recent years, as many companies integrate AI agents into their development processes, new challenges have emerged. A particular concern is the risk of privileged credentials remaining in AI prompt histories. Traditionally, developers had to input API keys and database credentials into chat interfaces for AI agents to perform infrastructure operations. This practice risked sensitive information being stored in external logs or training data.
Agent Kit addresses this challenge. AI agents directly utilize Keeper's CLI tools (Commander and Keeper Secrets Manager CLI) to execute commands within the developer's authenticated session. This allows for processing sensitive information without exposing it externally.
Craig Lurey, CTO and Co-founder of Keeper Security, stated:
"Agent Kit sets the standard for how AI agents should securely interact with critical enterprise data. Agents use encrypted CLI tools locally and process within the developer's own authenticated session. This mechanism enables AI-driven development while maintaining the principles of zero-knowledge."
The kit is optimized for modern development workflows and primarily offers the following features:
- Secure Secret Retrieval
Utilizes keeper-secrets skills to inject credentials into the local environment. Sensitive information is never displayed on the chat screen.
- Automated Vault Management
keeper-admin skills enable user and team management, as well as audit compliance, via Commander.
- Automated Configuration
keeper-setup skills are used to automatically build Keeper's security environment. This ensures a secure state from the initial stages of new projects.
Additionally, for teams using hosted or integrated AI environments, Model Context Protocol (MCP) server integration is available (compatible with Docker and Node configurations). This mechanism allows secrets to be retrieved via a running MCP server without using local CLI.
Even when AI agents use Keeper's CLI tools, all operations are subject to role-based access control and audit logs, just as they would be for human users.
Jeremy London, Director of AI and Threat Analysis at Keeper Security, commented:
"Development speed and security are not mutually exclusive. Agent Kit evolves AI from a mere conversational tool into a partner that protects security boundaries. By enabling secure use of credentials when needed, without directly handling them, it significantly reduces major risks in modern development environments."
Agent Kit is open-sourced under the Apache 2.0 license. It is available from the official Keeper Security GitHub, as well as the Claude and Cursor marketplaces.
About Keeper Security
Keeper Security is a rapidly growing cybersecurity software company protecting a wide range of businesses and users in over 150 countries. Built on zero-knowledge and zero-trust foundations, it is a pioneer in security designed for all IT environments. Its flagship product, KeeperPAM®, is an AI-powered, cloud-native platform that protects users, devices, and entire infrastructures from cyberattacks.
In the field of Privileged Access Management (PAM), Keeper Security is recognized in Gartner's "Magic Quadrant."
Agent Kit is a suite of features that directly integrates Keeper Secrets Manager and Commander with leading AI coding agents such as Claude Code, Cursor, Codex, and GitHub Copilot. It allows for the secure automation of security settings and management tasks.
In recent years, as many companies integrate AI agents into their development processes, new challenges have emerged. A particular concern is the risk of privileged credentials remaining in AI prompt histories. Traditionally, developers had to input API keys and database credentials into chat interfaces for AI agents to perform infrastructure operations. This practice risked sensitive information being stored in external logs or training data.
Agent Kit addresses this challenge. AI agents directly utilize Keeper's CLI tools (Commander and Keeper Secrets Manager CLI) to execute commands within the developer's authenticated session. This allows for processing sensitive information without exposing it externally.
Craig Lurey, CTO and Co-founder of Keeper Security, stated:
"Agent Kit sets the standard for how AI agents should securely interact with critical enterprise data. Agents use encrypted CLI tools locally and process within the developer's own authenticated session. This mechanism enables AI-driven development while maintaining the principles of zero-knowledge."
The kit is optimized for modern development workflows and primarily offers the following features:
- Secure Secret Retrieval
Utilizes keeper-secrets skills to inject credentials into the local environment. Sensitive information is never displayed on the chat screen.
- Automated Vault Management
keeper-admin skills enable user and team management, as well as audit compliance, via Commander.
- Automated Configuration
keeper-setup skills are used to automatically build Keeper's security environment. This ensures a secure state from the initial stages of new projects.
Additionally, for teams using hosted or integrated AI environments, Model Context Protocol (MCP) server integration is available (compatible with Docker and Node configurations). This mechanism allows secrets to be retrieved via a running MCP server without using local CLI.
Even when AI agents use Keeper's CLI tools, all operations are subject to role-based access control and audit logs, just as they would be for human users.
Jeremy London, Director of AI and Threat Analysis at Keeper Security, commented:
"Development speed and security are not mutually exclusive. Agent Kit evolves AI from a mere conversational tool into a partner that protects security boundaries. By enabling secure use of credentials when needed, without directly handling them, it significantly reduces major risks in modern development environments."
Agent Kit is open-sourced under the Apache 2.0 license. It is available from the official Keeper Security GitHub, as well as the Claude and Cursor marketplaces.
About Keeper Security
Keeper Security is a rapidly growing cybersecurity software company protecting a wide range of businesses and users in over 150 countries. Built on zero-knowledge and zero-trust foundations, it is a pioneer in security designed for all IT environments. Its flagship product, KeeperPAM®, is an AI-powered, cloud-native platform that protects users, devices, and entire infrastructures from cyberattacks.
In the field of Privileged Access Management (PAM), Keeper Security is recognized in Gartner's "Magic Quadrant."