[New Feature]: Make the pipeline verification logs more detailed when Policy-2.5 is triggered.

### Description of the new feature/enhancement

Directly based on https://github.com/microsoft/winget-pkgs/pull/302899.

Citing myself from https://github.com/microsoft/winget-pkgs/pull/290623#issuecomment-3383249099:

*"I admit it would be very convenient if the verification result JSON files actually stated which specific term it claimed it found, instead of just throwing an "NinjaCheetah.NUSGet triggered Algo Speak - Potential offensive speech" error with no further detailing of it."*

For the PRs related to NUSGet, they have for 7 weeks now bumped into that error no matter what. I've acquitted the tags, the ShortDescription, and almost certainly now also the fingerprint as being the cause, and there are no apparent indications in any form or way that I can see as to what *is* the cause then, which necessitates some overhaul of the pipeline logs.

### Proposed technical implementation details (optional)

In the possible outcomes of the pipeline verifications' JSON files, change `NinjaCheetah.NUSGet triggered Algo Speak - Potential offensive speech` → `NinjaCheetah.NUSGet triggered Algo Speak - Potential offensive speech in manifest. Triggered by text match: ____________`

Start by locating the pipeline verification logic that produces the JSON outcomes and the Policy-2.5 message. Trace how the NUSGet match is reported for a manifest, then define the output as done when it identifies the matched text while preserving the existing trigger context. No specific source file or test is named in the issue.

Written by the indexing model from the issue text.