fix(patch): report no-op edits clearly

[Cdo87]

Summary

• add explicit PatchResult noop/message fields
• report identical replace-mode edits as successful no-ops without writing
• treat V4A pure-context/no-effective-change updates as successful no-ops with guidance

Tests

• pytest -q tests/tools/test_patch_parser.py tests/tools/test_file_operations.py
• git diff --check

[teknium1]

Thanks for isolating the patch-tool behavior; the underlying premise is still present on current main (tools/fuzzy_match.py:69-70, reached by tools/file_operations.py:1578-1589).

Problems

• The no-op state does not survive the actual tool boundary. tools/file_tools.py:1848-1868 serializes the result, then every non-error result gets files_modified overwritten and is recorded as a write. With this PR, a no-op would therefore report noop: true while also claiming the target was modified.

Suggested changes

• Gate the successful-patch bookkeeping in tools/file_tools.py on not result_dict.get("noop"), preserving the no-op's empty modification result and avoiding stale/write-state updates.
• Add a regression test through patch_tool, not only ShellFileOperations / apply_v4a_operations.

This is an automated hermes-sweeper review.