bosun-config

Manage Bosun AI environment configuration through a central config.toml file.

3|Updated Feb 24, 2026
One-click install
npx skills add https://github.com/oddship/bosun --skill bosun-config
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: bosun-config
Source: https://github.com/oddship/bosun/tree/main/.pi/skills/bosun-config
Command: npx skills add https://github.com/oddship/bosun --skill bosun-config

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill helps you manage and understand the configuration of the Bosun AI environment, ensuring your AI agents are set up correctly with the desired models, sandbox rules, and environment variables.

Core Features & Use Cases

  • Model Tier Management: Define and update the AI models used by different agents.
  • Sandbox Customization: Configure process-level (bwrap) and tool-level restrictions for agent execution.
  • Environment Variable Control: Specify which environment variables are allowed within the sandbox.
  • Use Case: You need to switch Bosun to use a more powerful AI model for complex reasoning tasks. You would edit the config.toml file to update the model tier and then run just init to apply the changes.

Quick Start

Update the AI model used by the 'medium' tier by editing the config.toml file and then running just init.

Frequently Asked Questions about bosun-config

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

FAQPage Schema
How do I configure AI model tiers for my agent environment?

To configure AI model tiers, edit the `config.toml` file to update the model definitions for your agents, then run `just init` to apply the changes and regenerate the internal Pi configuration files.

How do I restrict filesystem access in an AI sandbox?

You restrict filesystem access by customizing sandbox restrictions in the `config.toml` file, which controls process-level (bwrap) and tool-level limitations for agent execution to manage the security posture.

How do I pass environment variables into an AI agent sandbox?

You pass environment variables by specifying the allowed variables in the `config.toml` file, which controls environment variable pass-through into the agent execution sandbox during the templated generation process.

What is the best way to switch an AI agent to a more powerful model for complex reasoning?

The best way to switch models is editing the `config.toml` file to update the target model tier, then running `just init` to apply the configuration changes to the Bosun AI environment for your agents.

Does configuring the Bosun AI environment require any external dependencies?

Configuring the Bosun AI environment does not require external dependencies, as the Skill relies on a central `config.toml` file and the `just init` command to manage settings and templated Pi configuration generation.

Why are my sandbox configuration changes not applying to the AI agent?

Sandbox configuration changes may not apply if you edited the `config.toml` file but did not run `just init` afterward, which is required to process the templated generation of internal Pi configuration files.