fix: reduce redundant context in apply_diff error messages

[KJ7LNW]

Context

When apply_diff fails to find a match, it previously returned an excessively verbose error message that included redundant information the model already had, wasting valuable context window space.

Implementation

This PR streamlines the error output by:

• Removing the redundant search content that was just sent by the model
• Removing the large original file content section that was noisy and context-consuming
• Keeping only the 'Best Match Found' section, which provides the most actionable information for debugging

The best match section shows exactly what was found in the file that was closest to the search pattern, enabling targeted corrections without re-reading the entire file.

How to Test

1. Use apply_diff with content that doesn't exactly match the file
2. Observe that the error message is now more concise but still provides the critical information needed for debugging

Get in Touch

Discord: KJ7LNW

Fixes #5795

---

Important

Streamlines apply_diff error messages by removing redundant content and updating version to 3.23.12.

• Behavior:
  • Streamlines error messages in apply_diff by removing redundant search content and large original file content sections in multi-file-search-replace.ts and multi-search-replace.ts.
  • Retains 'Best Match Found' section for actionable debugging information.
• Versioning:
  • Updates version to 3.23.12 in package.json to reflect changes.
• Misc:
  • Removes .changeset/v3.23.12.md file as part of cleanup.

^{This description was created by for e308c174f62c9a8775d3f5866f4dfedf3dd38465. You can customize this summary. It will automatically update as commits are pushed.}

[KJ7LNW]

rebased on v3.23.14

[daniel-lxs]

@roomote-agent Show me a before and after with this change, do not create a new PR or make any changes

[roomote]

I'll show you the before and after comparison of the apply_diff error message changes. Let me examine the code changes first.

[roomote]

Before and After Comparison: apply_diff Error Messages

This PR significantly reduces the verbosity of apply_diff error messages by removing redundant content while keeping the most actionable debugging information.

BEFORE (Previous Error Message):

When apply_diff failed to find a match, it would return a very verbose error like this:

No sufficiently similar match found (75% similar, needs 80%)

Debug Info:
- Similarity Score: 75%
- Required Threshold: 80%
- Search Range: starting at line 42
- Tried both standard and aggressive line number stripping
- Tip: Use the read_file tool to get the latest content of the file before attempting to use the apply_diff tool again, as the file content may have changed

Search Content:
function calculateTotal(items) {
    let sum = 0;
    for (const item of items) {
        sum += item.price;
    }
    return sum;
}

Best Match Found:
42 | function calculateTotal(items) {
43 |     let total = 0;
44 |     for (const item of items) {
45 |         total += item.price;
46 |     }
47 |     return total;
48 | }

Original Content:
35 | // Helper functions
36 | function validateInput(data) {
37 |     return data && typeof data === 'object';
38 | }
39 | 
40 | // Main calculation function
41 | function calculateTotal(items) {
42 |     let total = 0;
43 |     for (const item of items) {
44 |         total += item.price;
45 |     }
46 |     return total;
47 | }
48 | 
49 | function formatCurrency(amount) {
50 |     return `$${amount.toFixed(2)}`;
51 | }

AFTER (New Streamlined Error Message):

Now the same error is much more concise:

No sufficiently similar match found (75% similar, needs 80%)

Debug Info:
- Similarity Score: 75%
- Required Threshold: 80%
- Search Range: starting at line 42
- Tried both standard and aggressive line number stripping
- Tip: Use the read_file tool to get the latest content of the file before attempting to use the apply_diff tool again, as the file content may have changed

Best Match Found:
42 | function calculateTotal(items) {
43 |     let total = 0;
44 |     for (const item of items) {
45 |         total += item.price;
46 |     }
47 |     return total;
48 | }

Key Changes:

1. Removed Search Content section - This was redundant since the model just sent this content
2. Removed Original Content section - This large section was noisy and consumed valuable context window space
3. Kept Best Match Found section - This is the most actionable information for debugging, showing exactly what was found in the file

Benefits:

• Reduced context window usage - Saves valuable tokens for more important information
• Less noise - Focuses on what's actually helpful for debugging
• Still actionable - The 'Best Match Found' section shows exactly what needs to be corrected
• Maintains debugging capability - All essential information for fixing the issue is preserved

The error message is now much more focused and efficient while still providing the critical information needed to understand why the match failed and how to fix it.

[daniel-lxs]

LGTM

[mrubens]

It's not clear to me that the original content isn't valuable here. Is there a way we can demonstrate that it does more harm than good? I don't think that the best match on a failed diff is always the intended match. Until we have a little more evidence that this is positive I'm hesitant to make this change.

[KJ7LNW]

Run an eval?