Fix: _strip_ansi regex covers only a subset of CSI escape sequences (#52)

[tbrandenburg]

Summary

The _strip_ansi helper in tests/test_workflow_to_harness.py only stripped a narrow set of CSI escape-sequence final bytes, making the test suite fragile if Rich emits other valid sequences.

Root Cause

The helper hard-coded [mGKHFJ] instead of matching the full CSI final-byte range the helper name and docstring imply.

Changes

File	Change
tests/test_workflow_to_harness.py	Broadened _strip_ansi to strip the full CSI range
tests/test_workflow_to_harness.py	Added a regression test for cursor-movement CSI sequences

Testing

• Type check passes
• Unit tests pass
• Lint passes
• Verified the help-contract tests still pass via make qa

Validation

make qa && uv run --locked pytest tests/test_workflow_to_harness.py -k strip_ansi

Issue

Fixes #52

📋 Implementation Details

Implementation followed artifact:

/repo/.agents/issues/issue-52.md

Deviations from plan:

None

Automated implementation from investigation artifact

[tbrandenburg]

🔍 Automated Code Review

Summary

The fix now matches the broader CSI form expected by the investigation and the regression test covers both cursor movement and private-mode sequences.

Findings

✅ Strengths

• The helper now strips broader CSI sequences instead of the original narrow final-byte subset.
• The new regression test exercises both a cursor movement CSI code and a private-mode sequence.

⚠️ Suggestions (non-blocking)

• None.

🔒 Security

• No security concerns identified.

Checklist

• Fix addresses root cause from investigation
• Code follows codebase patterns
• Tests cover the change
• No obvious bugs introduced

Self-reviewed by Claude • Ready for human review

[tbrandenburg]

PR 55 Review

Recommendation

Approve.

Summary

The PR is focused and matches the issue report. The regex expansion in _strip_ansi() correctly covers the CSI forms exercised by Rich, and the new regression test is a good safeguard.

Findings

Low

1. The new .agents/issues/completed/issue-52.md artifact is tracked in the diff even though the repo’s established issue archive convention appears to live under .ghar/issues/completed/. It does not affect runtime behavior, but it is inconsistent with the existing documentation layout. If this file is intended to be a permanent archive, move it under the repo’s current archive path; otherwise remove it from the PR.

Validation

• make qa ✅
• ruff check ✅
• ruff format --check ✅
• python -m py_compile ✅
• pytest ✅ 138 passed
• uv build ✅

Notes

• No implementation-report deviations were documented beyond the intended regex/test update.
• No functional issues were found in tests/test_workflow_to_harness.py.