bagman

Secure AI agent credentials with storage, validation, and delegation controls.

111|20|Updated Feb 9, 2026
One-click install
npx skills add https://github.com/profbernardoj/everclaw --skill bagman-profbernardoj
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: bagman
Source: https://github.com/profbernardoj/everclaw/tree/main/security/bagman
Command: npx skills add https://github.com/profbernardoj/everclaw --skill bagman-profbernardoj

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill provides robust security patterns to protect sensitive credentials like private keys, API secrets, and wallet information when used by AI agents, preventing leaks and unauthorized access.

Core Features & Use Cases

  • Secure Storage: Integrates with secret managers like 1Password for runtime retrieval.
  • Leak Prevention: Implements output sanitization and input validation to block secret exposure and prompt injection attacks.
  • Delegated Access: Facilitates the use of session keys and on-chain delegation frameworks for fine-grained, time-bound permissions.
  • Use Case: An AI agent needs to manage cryptocurrency funds. This Skill ensures that the agent's private key is never exposed in logs or prompts, and any transaction it initiates is strictly limited by pre-defined rules and requires explicit confirmation.

Quick Start

Install the 1Password CLI and authenticate to begin managing your agent's secrets securely.

Frequently Asked Questions about bagman

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I secure AI agent private keys and prevent them from leaking in prompts?

To secure AI agent private keys, you can use patterns for secure storage, output sanitization, and leak prevention. This ensures sensitive credentials are never exposed in logs or prompts by enforcing strict access controls and validation layers.

What is on-chain delegation and how does it work for AI agents managing crypto wallets?

On-chain delegation for AI agents involves using session keys to grant fine-grained, time-bound permissions. This ensures that any transaction the agent initiates is strictly limited by pre-defined rules and requires explicit confirmation, safeguarding wallet credentials.

How do I stop prompt injection attacks from exposing API secrets in autonomous systems?

To stop prompt injection attacks from exposing API secrets, implement input validation and output sanitization layers. These validation mechanisms block unauthorized access and prevent secret exposure when AI agents process untrusted inputs.

Can I use 1Password for runtime retrieval of AI agent secrets?

Yes, you can integrate with secret managers like 1Password for runtime retrieval of AI agent secrets. This secure storage integration allows autonomous systems to fetch sensitive credentials dynamically without hardcoding them into the application.

What are the limitations of using session keys for AI agent wallet management?

Session keys for AI agent wallet management are limited by their fine-grained, time-bound permissions. While they prevent unauthorized transactions by enforcing explicit confirmation and pre-defined rules, they require integration with delegation frameworks to function properly.

Do I need the 1Password CLI to start managing AI agent secrets securely?

Yes, installing and authenticating the 1Password CLI is required to begin managing your agent's secrets securely. This environment setup provides the secure storage foundation necessary for robust leak prevention and credential protection.