Skip to content

TheSoftwareHouse/copilot-collections

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

319 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Copilot Collections

Opinionated GitHub Copilot setup for product discovery, implementation, and review.

Full documentation: https://copilot-collections.tsh.io/

This README is intentionally short. It covers the main workflows and the minimum setup needed so GitHub Copilot can use this repository effectively.

What Matters Most

You do not need to learn every agent, prompt, or skill up front.

For most teams, the main entry points are:

  • /tsh-analyze-materials for turning workshop inputs into structured tasks
  • /tsh-implement for research, planning, and implementation
  • /tsh-review for structured code review

The most important agents behind those workflows are:

  • tsh-business-analyst for discovery and backlog shaping
  • tsh-engineering-manager for orchestrating implementation work
  • tsh-code-reviewer for review and risk detection
  • tsh-copilot-orchestrator for creating or improving Copilot customizations

Everything else is supporting structure. Use the website for the full catalog, detailed flow descriptions, and deep documentation.

Requirements

  • GitHub Copilot Pro or higher
  • VS Code 1.109 or later
  • This repository available on disk so VS Code can read .github/prompts, .github/agents, .github/skills, and .vscode/mcp.json

Recommended Installation: Ask Copilot To Configure Itself

The preferred setup is to let GitHub Copilot update the required user-level configuration for you.

1. Clone this repository

cd ~/projects
git clone <this-repo-url> copilot-configuration

If you clone it somewhere else, replace the path in the prompt below.

2. Ask Copilot to configure your machine

Open any VS Code workspace, open Copilot Chat, and paste this prompt:

Configure GitHub Copilot on this machine to use the Copilot Collections repository at /Users/adampolak/projects/copilot-configuration.

Update my VS Code User Settings so Copilot loads:
- prompts from /Users/adampolak/projects/copilot-configuration/.github/prompts
- agents from /Users/adampolak/projects/copilot-configuration/.github/agents
- skills from /Users/adampolak/projects/copilot-configuration/.github/skills

Enable the required Copilot settings for custom agents and skills.

Then open my user MCP configuration and copy in the MCP server template from /Users/adampolak/projects/copilot-configuration/.vscode/mcp.json.

Do not change project source files. Only modify user-level Copilot and MCP configuration.

3. If Copilot asks what settings to add, use these

These are the minimum VS Code user settings this repository expects:

{
  "chat.promptFilesLocations": {
    "/Users/adampolak/projects/copilot-configuration/.github/prompts": true
  },
  "chat.agentFilesLocations": {
    "/Users/adampolak/projects/copilot-configuration/.github/agents": true
  },
  "chat.agentSkillsLocations": {
    "/Users/adampolak/projects/copilot-configuration/.github/skills": true
  },
  "chat.useAgentSkills": true,
  "github.copilot.chat.searchSubagent.enabled": true,
  "chat.experimental.useSkillAdherencePrompt": true,
  "chat.customAgentInSubagent.enabled": true,
  "github.copilot.chat.agentCustomizationSkill.enabled": true
}

For MCP, copy the template from .vscode/mcp.json into your user MCP configuration using MCP: Open User Configuration.

Manual Fallback

If Copilot cannot complete the setup automatically:

  1. Open Preferences: Open User Settings (JSON) and add the settings shown above.
  2. Run MCP: Open User Configuration.
  3. Copy the contents of .vscode/mcp.json into your user MCP config.

First Commands To Try

Once configured, open your target project and start with one of these:

  • /tsh-implement <task or Jira ID>
  • /tsh-review <task or Jira ID>
  • /tsh-review-ui
  • /tsh-analyze-materials <transcript, notes, links, or workshop assets>

If you want to extend the system itself, use:

  • /tsh-create-custom-agent
  • /tsh-create-custom-skill
  • /tsh-create-custom-prompt
  • /tsh-create-custom-instructions

Learn More

License

MIT

About

Opinionated AI-enabled workflows for product engineering

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors