ayjnt-tools

Name: ayjnt-tools
Availability: InStock
Author: Northclock

Bridge workerd and Bun host processes to create permissioned file and command tools.

3|1|Updated Apr 14, 2026

Search Tags:#automation #security #agents #bun #cloudflare #tools

Ready to run

npx skills add https://github.com/northclock/ayjnt --skill ayjnt-tools

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: ayjnt-tools
Source: https://github.com/northclock/ayjnt/tree/main/.claude/skills/ayjnt-tools
Command: npx skills add https://github.com/northclock/ayjnt --skill ayjnt-tools

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill solves the challenge of integrating local host capabilities, such as file system access or shell commands, into Cloudflare-based agent runtimes while maintaining security and deployment integrity.

Core Features & Use Cases

Dual-Runtime Support: Seamlessly define tools that run either in the workerd environment or on the Bun host process.
Permissioned Execution: Enforce strict side-effect controls (read, write, exec) for host-side tools to prevent unauthorized system access.
Use Case: Build a coding agent that can read project files and execute shell commands locally while keeping the core agent logic deployable as a standard Cloudflare worker.

Quick Start

Add a new host tool to your agent by creating a tools.host.ts file in your agent route directory and defining your function with the hostTool wrapper.

Frequently Asked Questions about ayjnt-tools

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

FAQPage Schema

How do I give my Cloudflare agent access to local file system and shell commands?▼

To give a Cloudflare agent local file system access, you use a skill that bridges workerd and Bun host processes. This allows you to define permissioned tools for file interaction and command execution while keeping the agent deployable as a standard worker.

How do I enforce security and permissions for agent tools that execute local shell commands?▼

To enforce security for agent tools executing local shell commands, you use side-effect declarations and path confinement validation. This permissioned execution model prevents unauthorized system access by strictly controlling read, write, and exec operations.

Can I run agent tools on a Bun host process while keeping the core logic in workerd?▼

Yes, you can run agent tools on a Bun host process while keeping core logic in workerd. This dual-runtime support allows you to seamlessly define tools that operate in either environment, facilitating local host-side capabilities integration.

What is the best way to add per-route model tools to an agent without compromising deployment integrity?▼

The best way to add per-route model tools without compromising deployment is by using a framework that bridges workerd and Bun processes. This approach maintains security through side-effect controls and path confinement, ensuring your worker remains standard and deployable.

How do I define a host tool for my agent using the ayjnt framework?▼

To define a host tool in the ayjnt framework, you create a tools.host.ts file in your agent route directory. You then define your function using the hostTool wrapper to enable local execution.

Does integrating local host capabilities into a Cloudflare agent runtime require special side-effect declarations?▼

Yes, integrating local host capabilities requires special side-effect declarations. You must enforce strict side-effect controls for read, write, and exec operations to prevent unauthorized system access and ensure secure tool invocation.