fix(powershell): ensure UTF-8 templates are written without BOM

[Nimraakram22]

Description
This PR fixes an issue where PowerShell 5.1 (default on Windows) writes files like CLAUDE.md with a UTF-8 Byte Order Mark (BOM). This invisible prefix causes parsing errors in modern tools like Claude Code and other Unix-style CLI utilities.

Changes
Stripped BOM from existing templates: Found and removed the BOM from all .md files in the templates/ and presets/ directories.

Hardened PowerShell logic: Replaced Copy-Item with a .NET based write method ([System.IO.File]::WriteAllText) in setup-plan.ps1 and create-new-feature.ps1 to force "UTF-8 No BOM" output, regardless of the source template's encoding.

Verification
Verified using the file command in MINGW64 to ensure output files now show as UTF-8 Unicode text instead of UTF-8 Unicode (with BOM) text.

[Copilot]

Pull request overview

This PR updates the PowerShell feature/setup scripts to ensure generated markdown files are written as UTF-8 without BOM (avoiding BOM-related parsing issues in Unix-style tooling).

Changes:

• Replaced Copy-Item template copying with a read + [System.IO.File]::WriteAllText(...) flow to force UTF-8 no BOM output.
• Applied the same UTF-8 no BOM write approach in both setup-plan.ps1 and create-new-feature.ps1.

Show a summary per file

File	Description
scripts/powershell/setup-plan.ps1	Writes plan.md from the template using UTF-8 no BOM instead of Copy-Item.
scripts/powershell/create-new-feature.ps1	Writes spec.md from the template using UTF-8 no BOM instead of Copy-Item.

Copilot's findings

Tip

Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Comments suppressed due to low confidence (2)

scripts/powershell/setup-plan.ps1:39

• Variable $Utf8NoBom is PascalCase while other local variables in this file are lower camelCase (e.g., $paths, $template, $content). Consider renaming to match the local variable naming convention (e.g., $utf8NoBom).

    $Utf8NoBom = New-Object System.Text.UTF8Encoding($false)

scripts/powershell/create-new-feature.ps1:355

• Local variables in this script use lower camelCase (e.g., $featureDesc, $branchName), but $Utf8NoBom is PascalCase. Consider renaming to match the file's convention (e.g., $utf8NoBom).

            $Utf8NoBom = New-Object System.Text.UTF8Encoding($false)

• Files reviewed: 2/2 changed files
• Comments generated: 3

[mnriem]

Please address Copilot feedback. If not appicable, please explain why

[Nimraakram22]

@mnriem Done! I have updated both setup-plan.ps1 and create-new-feature.ps1 to:

Use -Encoding UTF8 when reading templates to ensure character integrity.

Use lower camelCase for the $utf8NoBom variable.

Fix the indentation of the .NET write method calls.

[Copilot]

Copilot's findings

Tip

Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Comments suppressed due to low confidence (2)

scripts/powershell/create-new-feature.ps1:355

• The PR description says the write is hardened "regardless of the source template's encoding", but Get-Content -Encoding UTF8 forces UTF-8 decoding. If a resolved template is UTF-16/UTF-32 (common for Windows-authored files) this will corrupt content. Consider reading via a BOM-detecting API (e.g., [System.IO.File]::ReadAllText($template) / StreamReader with BOM detection) and then writing with UTF8Encoding($false).

            $content = Get-Content -Raw -Encoding UTF8 -Path $template
            $utf8NoBom = New-Object System.Text.UTF8Encoding($false)

scripts/powershell/setup-plan.ps1:41

• Get-Content -Encoding UTF8 forces UTF-8 decoding, so this doesn't fully meet the stated goal of being robust to template encoding (e.g., UTF-16 templates with a BOM would be mis-decoded). Prefer a BOM-detecting read (such as [System.IO.File]::ReadAllText($template)) before writing out with UTF8Encoding($false).

    $content = Get-Content -Raw -Encoding UTF8 -Path $template
    $utf8NoBom = New-Object System.Text.UTF8Encoding($false)
    [System.IO.File]::WriteAllText($paths.IMPL_PLAN, $content, $utf8NoBom)

• Files reviewed: 2/2 changed files
• Comments generated: 2

[mnriem]

Please address Copilot feedback. If not applicable please explain why

[Nimraakram22]

@mnriem "Fixed! I have updated the scripts based on Copilot's feedback:

Switched to [System.IO.File]::ReadAllText() for more robust template encoding detection.

Corrected the variable name from $targetPath to $specFile in create-new-feature.ps1.

Cleaned up the duplicate comments and indentation.

Ready for another look!"

[Copilot]

Copilot's findings

Tip

Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

• Files reviewed: 2/2 changed files
• Comments generated: 0 new