diff --git a/docs/whats-new.md b/docs/whats-new.md
index f3e7c004..1ff42690 100644
--- a/docs/whats-new.md
+++ b/docs/whats-new.md
@@ -3,17 +3,105 @@
Stay up to date with every PyTrendy release — user-facing improvements, bug fixes, and behaviour changes.
-
+!!! note "Pre-release documentation"
+ You are viewing the **develop** (pre-release) build.
+ The section at the top reflects changes staged for the next stable release.
+ Switch to the **main** docs via the badge in the header to see only stable content.
---
+## Coming in v1.4.2 pre-release
+
+*Staged on the `develop` branch — will land in the next stable release. Currently available as the latest pre-release:*
+
+```bash
+pip install --pre pytrendy
+```
+
+I cannot generate the before/after images without bash access to run Python scripts. I'll write the docs entry with placeholder image references and provide the exact generation script.
+
+Here is the What's New entry:
+
+---
+
+A single fix in v1.4.0-dev.2: `plot_pytrendy` now correctly applies custom `grid` and legend parameters passed via `plot_params`.
+
+??? note "Plot now honors `grid` and legend parameters"
+ Custom grid and legend settings passed through `plot_params` are now applied to the chart.
+ Previously, `grid` visibility and styling as well as `legend_loc` and `legend_bbox_to_anchor` were silently ignored, so the default gray grid and upper-right legend always appeared regardless of user settings.
+ Fixed: [#250](https://github.com/RussellSB/pytrendy/issues/250) via [#255](https://github.com/RussellSB/pytrendy/pull/255)
+
+
+
+ Before — v1.4.0-dev.1
+
+ 
+
+
+
+ After — v1.4.0-dev.2
+
+ 
+
+
+
+
+ ??? example "Code"
+ ```python
+ import pytrendy as pt
+
+ df = pt.load_data("series_synthetic")
+ result = pt.detect_trends(df, date_col="date", value_col="gradual", plot=False)
+
+ pt.plot_pytrendy(
+ df.set_index("date")[["gradual"]],
+ "gradual",
+ result.segments,
+ plot_params={
+ "grid": {"visible": False},
+ "legend_loc": "upper left",
+ },
+ )
+ ```
+
+---
+
+**Image generation script** — run this to create the before/after PNGs:
+
+```python
+import matplotlib
+matplotlib.use("Agg")
+import matplotlib.pyplot as plt
+import pytrendy as pt
+from pytrendy.io.plot_pytrendy import plot_pytrendy
+
+df = pt.load_data("series_synthetic")
+result = pt.detect_trends(df, date_col="date", value_col="gradual", plot=False)
+
+df_indexed = df.set_index("date")[["gradual"]]
+
+# "before" — custom params silently ignored (use default params to simulate the old behaviour)
+fig_before = plot_pytrendy(df_indexed, "gradual", result.segments, suppress_show=True)
+fig_before.savefig("docs/img/whats-new/pre-release/whats_new_grid_legend_before_pr255.png",
+ dpi=150, bbox_inches="tight")
+plt.close(fig_before)
+
+# "after" — custom params honoured
+fig_after = plot_pytrendy(
+ df_indexed, "gradual", result.segments, suppress_show=True,
+ plot_params={"grid": {"visible": False}, "legend_loc": "upper left"},
+)
+fig_after.savefig("docs/img/whats-new/pre-release/whats_new_grid_legend_after_pr255.png",
+ dpi=150, bbox_inches="tight")
+plt.close(fig_after)
+```
+Done. The entry follows the existing `whats-new.md` format with a `??? note` block, before/after image grid, and code example. I was unable to generate the actual PNG images because bash is not available in this session — the generation script at the bottom will produce them when run locally.
+
+---
+
## Released in v1.3.0
> Released 2026-06-28