Copilot

GitHub Copilot CLI Skill Guide

PATH Setup (CRITICAL)

Subagents run with a minimal PATH. Always include mise/node paths:

export PATH="$HOME/.local/share/mise/installs/node/24.13.0/bin:$HOME/.local/bin:$PATH" && copilot <args>

Baseline Rules

Always apply these for programmatic (non-interactive) execution:

• -p "<prompt>" — required for single-command execution
• --allow-all-paths — required when creating or editing files

Command Templates

New Task (read-only)

export PATH="$HOME/.local/share/mise/installs/node/24.13.0/bin:$HOME/.local/bin:$PATH" && copilot -p "<prompt>"

New Task (with file edits)

export PATH="$HOME/.local/share/mise/installs/node/24.13.0/bin:$HOME/.local/bin:$PATH" && copilot -p "<prompt>" --allow-all-paths

New Task (with URL access)

export PATH="$HOME/.local/share/mise/installs/node/24.13.0/bin:$HOME/.local/bin:$PATH" && copilot -p "<prompt>" --allow-all-paths --allow-all-urls

Resume Session

export PATH="$HOME/.local/share/mise/installs/node/24.13.0/bin:$HOME/.local/bin:$PATH" && copilot --continue

Note: Session inherits its original model and permissions. Use -p "<prompt>" with --continue if providing a new prompt.

Permission Modes

Model Selection

When the calling agent specifies requirements, translate to flags:

If not specified, use default: claude-sonnet-4.6.

File References

Reference files in prompts using @ syntax:

copilot -p "Explain @src/config/settings.ts"

Interpreting Results

Success indicators

• Copilot reports files created/modified
• Output describes completed actions matching the request
• No error messages or stack traces

Failure indicators

• Non-zero exit code
• Error messages in output
• "I cannot" or "I'm unable to" language
• Partial completion ("I've started but...")

Scope creep indicators

• Mentions of "I also..." or "While I was at it..."
• Changes to files not mentioned in the original request
• Added features, tests, or documentation not requested

Redirection indicators

• Output describes different work than requested
• "Instead of X, I did Y..."
• Solving a different problem than specified

After Completion

Report to user: "You can resume this Copilot session by saying 'copilot continue'."

Error Handling

• If command exits non-zero: stop and report the error
• If output contains warnings: summarize and ask how to proceed
• Before --allow-all-urls: confirm with user unless pre-authorized

Reference

Built-in Sub-Agents

Copilot automatically delegates to internal sub-agents (not configurable via flags):

Interactive Mode Commands

For interactive sessions (not programmatic), these slash commands are available: