Scripting the GitHub Copilot CLI - Deep Dive
summary
Dives into the Copilot CLI workflow: runs Copilot from the terminal with -p (prompt) and allow tools to emit standard out, uses command substitution and globbed file preloads to pass real repo context, converts diffs into auto commit messages, and manages sessions with continue/resume for chained work. Shows ZSH aliases to shorten commands, builds markdown-based agents and a tool called MDFlow, and wires agents into a pre-commit hook that blocks commits when tests are missing.
AI-written summary · about
chapters (16)
- 0:00Intro and Overview of Copilot in the Terminal
- 0:49Running Copilot with Prompt Mode and Standard Output
- 1:30Preloading Context and Using Markdown for AI
- 2:22Using Diffs and Automating Commit Messages
- 3:03Key Flags: -p, Allow Tools, and Workflow Security
- 3:34Demo: Running Copilot in the Terminal
- 4:10Setting Up ZSH Aliases for Copilot Commands
- 5:54Interactive Mode vs Prompt Mode Explained
- 8:49Combining AI with Command Substitution and GitHub CLI
- 11:30Continue and Resume Sessions for Workflow Efficiency
- 13:41Building Custom Agents with Markdown
- 16:36Automating Pre-Commit Hooks with AI Agents
- 18:15Introducing MDFlow: Markdown That Thinks
- 20:49Running Markdown Files as Intelligent Agents
- 23:40Advanced Features: Piping, Templates, and Conditional Logic
- 25:55Wrap-Up and Final Thoughts
related talks
48mThe Terminal Strikes Back, Agentic Engineering using the CLI
40mUnlocking Productivity with GitHub Copilot Best Practices and Tips
45mAgentic AI for Maintainers: Getting the Most From GitHub Copilot Enterprise
01h 04mCollaborating with Agents in your Software Dev Workflow
01h 04mThe Agent Awakens: Collaborative Development with Copilot
01h 20mGithub Copilot - The World's Most Widely Adopted AI Developer Tool
01h 05mThe Next Generation of Developer Productivity with GitHub Copilot & Visual Studio
01h 00mLevel up with GitHub Copilot: using AI to learn, code, and build