[FEATURE] - Add GitHub Copilot target

[momostallion]

Feature Description

GitHub Copilot supports custom instructions, prompt files, and agents via the .github directory. This is used by both VS Code and Visual Studio.

Proposed mapping

Lola concept	Copilot path	Format
Skills	.github/instructions/.instructions.md	Markdown with optional applyTo frontmatter
Commands	.github/prompts/.prompt.md	Markdown prompt file
Agents	.github/agents/.agent.md	Markdown with YAML frontmatter

Notes

• Copilot also reads .github/copilot-instructions.md as a global instruction file — could optionally append a skills index there (similar to how gemini-cli and opencode use managed sections in GEMINI.md / AGENTS.md)
• applyTo in frontmatter lets skills scope to specific file patterns (e.g., **/*.tf for a Terraform skill)
• Works for both VS Code and Visual Studio (same file paths)

References

• VS Code docs: Custom instructions
• Visual Studio docs: Customize chat responses

Problem or Use Case

GitHub Copilot (VS Code + Visual Studio) is currently the most widely adopted AI coding assistant but has no lola target. Users who work in VS Code/Visual Studio must manually copy skill files or maintain parallel configurations. Adding a native target lets lola manage Copilot context the same way it does for Claude Code, Cursor, and OpenCode.

Category

New target assistant support

Proposed Solution

Add a copilot target to targets.py that maps lola's module structure to GitHub Copilot's .github directory convention:

• Skills → .github/instructions/.instructions.md (with applyTo frontmatter preserved)
• Commands → .github/prompts/.prompt.md
• Agents → .github/agents/.agent.md

copilot would be added to the TARGETS registry so it appears in the interactive assistant selector during lola install and can be specified via lola install -a copilot.

Alternatives Considered

Users can manually symlink or copy lola-installed skill files (e.g., from .claude/skills/) into .github/instructions/, but this breaks the "install once, target many" workflow and requires manual upkeep on every lola update.

Complexity Estimate

Simple (good first issue)

Additional Context

No response

[momostallion]

happy to implement after 125 wraps up

[momostallion]

I went ahead and implemented this in PR #154. I found out that I'm going to lose access to contribute to OS on GH temporarily at my org so wanted to get the PR out if there were any changes needed.