fix: correct formatting of <thinking> tags in prompts for improved output clarity

Author: hannesrudolph

docs/update-notes/v3.28.mdx

@@ -94,7 +94,7 @@ This gives teams a higher-capacity OpenAI option without extra configuration.[#8
 * **Reset logs you out of Roo Code Cloud**: Reset now also logs you out of Roo Code Cloud for a truly clean slate; the reset still completes even if logout fails. ([#8312](https://github.com/RooCodeInc/Roo-Code/pull/8312))
 * **Measure engagement with upsell messages**: Telemetry for DismissibleUpsell tracks dismiss vs click (upsellId only) to tune messaging and reduce noise; no user-facing behavior change. ([#8309](https://github.com/RooCodeInc/Roo-Code/pull/8309))
 * **Welcome screen Roo provider experiment**: Surfaces the Roo provider on the welcome screen (localized, with click telemetry) to speed up setup. ([#8317](https://github.com/RooCodeInc/Roo-Code/pull/8317))
-* **Cleaner prompts (no “thinking” tags)**: Removes <thinking> tags for cleaner output, fewer tokens, and better model compatibility; preserves the plain‑language rule to confirm tool success. ([#8319](https://github.com/RooCodeInc/Roo-Code/pull/8319))
+* **Cleaner prompts (no “thinking” tags)**: Removes `<thinking>` tags for cleaner output, fewer tokens, and better model compatibility; preserves the plain‑language rule to confirm tool success. ([#8319](https://github.com/RooCodeInc/Roo-Code/pull/8319))
 * **Simpler Requesty model refresh**: Removes an unnecessary warning when refreshing the Requesty models list so you can fetch new policies without leaving the screen. ([#7710](https://github.com/RooCodeInc/Roo-Code/pull/7710))
 
 ## Bug Fixes