fix: replace non-ASCII chars in PowerShell scripts for PS 5.1 compat

[mnriem]

Summary

Fixes #2680

PowerShell 5.1 (built-in on Windows) defaults to the system's legacy encoding when reading .ps1 files without a BOM. Multi-byte UTF-8 sequences for characters like ✓ (U+2713) are misinterpreted as individual bytes, one of which gets parsed as a quote character, causing TerminatorExpectedAtEndOfString parse errors.

Changes

PowerShell script fixes (5 files)

• Replace ✓ → [OK] and ✗ → [FAIL] in string literals (common.ps1, initialize-repo.ps1)
• Replace — (em-dash) → -- in comments across all affected .ps1 files

Test fixes (2 files)

• tests/integrations/test_cli.py — apply strip_ansi() to capsys output before asserting against Rich-formatted text
• tests/integrations/test_integration_subcommand.py — import and apply strip_ansi() to CLI output assertions affected by Rich ANSI escape codes

Regression tests (1 new file)

• tests/test_ps1_encoding.py — 10 parametrized tests scanning every shipped .ps1 file byte-by-byte, asserting all content is ASCII-only (≤ 0x7F). Includes a discovery sanity check.

Testing

• Full suite: 3018 passed, 1 skipped (pre-existing Windows-only test), 0 failed
• Negative case verified: injecting ✓ back into common.ps1 causes immediate test failure

[Copilot]

Pull request overview

This PR addresses PowerShell 5.1 script parsing failures on Windows caused by non-ASCII characters being mis-decoded when .ps1 files are read without a BOM, and adds regression coverage to prevent reintroduction.

Changes:

• Replaced non-ASCII symbols (e.g., ✓, ✗, em-dash) in shipped PowerShell scripts to ensure ASCII-only content.
• Updated CLI/integration tests to strip Rich ANSI escape codes before making string assertions.
• Added a regression test suite that scans shipped .ps1 files and fails on any non-ASCII bytes.

Show a summary per file

File	Description
tests/test_ps1_encoding.py	Adds regression tests enforcing ASCII-only bytes in shipped PowerShell scripts.
tests/integrations/test_integration_subcommand.py	Normalizes CLI output by stripping ANSI codes before parsing/asserting table and messages.
tests/integrations/test_cli.py	Strips ANSI codes from captured output before assertions to avoid Rich formatting affecting tests.
scripts/powershell/create-new-feature.ps1	Replaces non-ASCII em-dash in a comment for PS 5.1-safe encoding.
scripts/powershell/common.ps1	Replaces Unicode status glyphs and em-dash usage to keep script content ASCII-only.
extensions/git/scripts/powershell/initialize-repo.ps1	Replaces Unicode checkmark output and em-dash in comment for PS 5.1 compatibility.
extensions/git/scripts/powershell/git-common.ps1	Replaces non-ASCII em-dash in comment to keep the file ASCII-only.
extensions/git/scripts/powershell/auto-commit.ps1	Replaces non-ASCII em-dash in comment to keep the file ASCII-only.

Copilot's findings

Tip

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

• Files reviewed: 8/8 changed files
• Comments generated: 2

[Copilot]

Copilot's findings

• Files reviewed: 8/8 changed files
• Comments generated: 1

[mnriem]

Closing as superseded by PR #2709 (c7e0cac), which merged the same fix to main with two additional improvements from review feedback:

1. glob("*.ps1") → rglob("*.ps1") for recursive script discovery in regression tests
2. Updated speckit.git.initialize.md docs to reflect the [OK] output change

Issue #2680 has been closed as completed.