diff --git a/flow/designs/README.md b/flow/designs/README.md
new file mode 100644
index 0000000000..c2b471eff6
--- /dev/null
+++ b/flow/designs/README.md
@@ -0,0 +1,57 @@
+# ORFS designs
+
+## Findings: how accurate are early-stage WNS estimates?
+
+Reading the committed `rules-base.json` baselines, normalized by each design's clock
+period (parsed from its `.sdc`), the picture across the 67 designs / 8 PDKs that expose
+`cts` and `globalroute` slack is:
+
+- **Global route usually tightens the estimate.** For most PDKs the mean absolute error
+  drops from `cts` to `globalroute` — dramatically for `sky130hs` (10.5% → 1.9%) and
+  `gt2n` (3.3% → 0.0%), and clearly for `gf12` (2.2% → 1.1%) and `ihp-sg13g2`
+  (0.6% → 0.0%). `ihp-sg13g2` and `gf180` are already accurate at `cts`.
+
+- **`cts` is biased optimistic; `globalroute` often overshoots into pessimism.** Every
+  PDK's `cts` bias is ≥ 0 (cts reports more slack than the design finally closes with),
+  whereas `globalroute` bias flips negative for `sky130hd` (−3.5%), `sky130hs` (−1.9%),
+  `gf12` (−1.1%) and `nangate45` (−0.5%). Global route tends to *over-correct*.
+
+- **`sky130hd` is the exception where routing makes the estimate worse**, not better:
+  `globalroute` MAE (3.5%) exceeds `cts` MAE (2.9%), and it is consistently pessimistic.
+
+- **Outliers are design-specific, not PDK-wide.** `sky130hs/gcd` has `cts` +45.9%
+  (wildly optimistic, fully corrected by `globalroute`), and `asap7/swerv_wrapper` is
+  +14.9% optimistic at *both* stages — the cases most likely to mislead an early-stage
+  go/no-go decision.
+
+Practical reading: `cts` slack is a usable optimistic rank-ordering; `globalroute` is the
+first estimate within a few % of final for most PDKs, but on `sky130hd` (and for specific
+designs elsewhere) even `globalroute` can be off by 3–10% of the clock period. This is the
+design-level companion to the per-net GRT-vs-RCX divergence in
+[`flow/docs/rcx`](../docs/rcx/README.md) (PR #4302). Per-PDK design breakdowns:
+[asap7](asap7/README.md), [nangate45](nangate45/README.md), [sky130hd](sky130hd/README.md),
+[sky130hs](sky130hs/README.md), [gf12](gf12/README.md), [gf180](gf180/README.md),
+[gt2n](gt2n/README.md), [ihp-sg13g2](ihp-sg13g2/README.md).
+
+<!-- BEGIN WNS-ACCURACY (generated by flow/util/plot_wns.py) -->
+## WNS estimate accuracy across PDKs
+
+How closely the earlier-stage worst-slack estimates (`cts`, `globalroute`) match the final (`finish`) WNS, per design, normalized by that design's clock period so PDKs with different timing units are comparable. Error is `(stage − finish) / clock_period`; **positive = optimistic** (the stage reported more slack than the design actually closes with), negative = pessimistic. Clock period is parsed from each design's `.sdc`; designs whose period could not be parsed are omitted.
+
+![WNS estimate accuracy by stage, across PDKs](wns_accuracy.png)
+
+Mean absolute error (MAE) and mean signed error (bias), in % of clock period:
+
+| PDK | designs | cts MAE | cts bias | grt MAE | grt bias | worst (design) |
+| --- | ---: | ---: | ---: | ---: | ---: | --- |
+| asap7 | 16 | 2.8% | +1.5% | 2.9% | +1.1% | +14.9% (swerv_wrapper globalroute) |
+| gf12 | 9 | 2.2% | -2.2% | 1.1% | -1.1% | -14.2% (jpeg cts) |
+| gf180 | 5 | 1.0% | +1.0% | 0.4% | +0.3% | +4.3% (aes cts) |
+| gt2n | 3 | 3.3% | +3.3% | 0.0% | +0.0% | +10.0% (aes cts) |
+| ihp-sg13g2 | 7 | 0.6% | +0.6% | 0.0% | +0.0% | +4.5% (spi cts) |
+| nangate45 | 15 | 0.8% | +0.6% | 1.0% | -0.5% | -3.2% (black_parrot globalroute) |
+| sky130hd | 7 | 2.9% | +2.0% | 3.5% | -3.5% | -10.0% (gcd globalroute) |
+| sky130hs | 5 | 10.5% | +10.5% | 1.9% | -1.9% | +45.9% (gcd cts) |
+
+_Generated by `flow/util/plot_wns.py`; regenerate with `python3 flow/util/plot_wns.py`._
+<!-- END WNS-ACCURACY -->
diff --git a/flow/designs/asap7/README.md b/flow/designs/asap7/README.md
new file mode 100644
index 0000000000..901603f308
--- /dev/null
+++ b/flow/designs/asap7/README.md
@@ -0,0 +1,68 @@
+# asap7 designs
+
+## Findings: worst negative slack (WNS) across asap7
+
+These notes read the committed `rules-base.json` baselines — the worst setup slack each
+design reaches at clock-tree synthesis (`cts`), global route (`globalroute`) and `finish`.
+The plot and table below are regenerated from that data by
+[`flow/util/plot_wns.py`](../../util/plot_wns.py); no flow run is needed, so the numbers
+are exactly what CI checks against.
+
+What stands out:
+
+- **Every asap7 design closes with negative setup slack.** All 18 baselines have
+  `finish` WNS < 0, so the committed asap7 constraints target clocks the flow does not
+  actually meet — these baselines track *how far short* each design lands, not timing
+  closure. `swerv_wrapper` (≈ −318) and `mock-alu` (≈ −300) are the extreme cases, an
+  order of magnitude worse than the cluster around −15 … −50 (`aes*`, `mock-cpu`,
+  `uart`, `jpeg`).
+
+- **Worst slack is often not stable across stages.** Some designs pin their critical
+  path early and barely move (`cva6`, `jpeg`, `mock-cpu`: `cts ≈ globalroute ≈ finish`).
+  Others move a lot: `swerv_wrapper` degrades from −80 at cts/globalroute to −318 at
+  finish, and `riscv32i`/`riscv32i-mock-sram` slip from ≈ −47 (cts) to ≈ −81 (finish) —
+  routing and final extraction make the path materially worse than CTS predicted.
+
+- **Global route is sometimes more pessimistic than finish.** For `gcd`
+  (cts −85.9 → globalroute −110 → finish −96.7) and `aes-block`
+  (cts −84 → globalroute −31 → finish −57) the worst-slack estimate swings between
+  stages rather than monotonically worsening. This is the same GRT-vs-post-route
+  estimate gap explored in [`flow/docs/rcx`](../../docs/rcx/README.md) (PR #4302), here
+  visible at the design level rather than per net.
+
+The takeaway for anyone using cts-stage slack as a proxy for the final result: for asap7
+it is a usable rank ordering but not a reliable magnitude — several designs move tens of
+units (and `swerv_wrapper` hundreds) between cts and finish.
+
+<!-- BEGIN WNS (generated by flow/util/plot_wns.py) -->
+## WNS
+
+Worst setup slack per design at three flow stages — clock-tree synthesis (`cts`), global route (`globalroute`) and `finish` — read from each design's `rules-base.json`. Negative means setup timing is not met. Values are in this PDK's native timing unit (ps for `asap7`, ns for most others), so they are comparable within this PDK but not across PDKs.
+
+The bar is the `finish` slack; the markers show the `cts` and `globalroute` slack for the same design, so stage-to-stage movement is visible.
+
+![WNS by design — asap7](wns.png)
+
+| design | cts | globalroute | finish |
+| --- | ---: | ---: | ---: |
+| swerv_wrapper | -80 | -80 | -318 |
+| mock-alu | -289 | -303 | -300 |
+| gcd | -85.9 | -110 | -96.7 |
+| gcd-ccs | -101 | -93.7 | -96.4 |
+| ethmac | -88.6 | -103 | -93.3 |
+| riscv32i-mock-sram | -47.5 | -52.8 | -81.3 |
+| riscv32i | -47.5 | -49.8 | -81.2 |
+| jpeg_lvt | -30 | -30 | -63.9 |
+| aes-block | -84.1 | -31.1 | -56.6 |
+| ibex | -79.4 | -74.3 | -52.5 |
+| cva6 | -50 | -50 | -50 |
+| uart | -47.6 | -58.7 | -49.1 |
+| jpeg | -34 | -34 | -34 |
+| aes_lvt | -18 | -18 | -26.1 |
+| aes | -28.9 | -28 | -24.2 |
+| aes-mbff | -20.8 | -19 | -20.8 |
+| mock-cpu | -16.6 | -16.6 | -16.6 |
+| ethmac_lvt | -19 | -29.5 | -15.2 |
+
+_Generated by `flow/util/plot_wns.py` from `rules-base.json`; regenerate with `python3 flow/util/plot_wns.py`._
+<!-- END WNS -->
diff --git a/flow/designs/asap7/wns.png b/flow/designs/asap7/wns.png
new file mode 100644
index 0000000000..2a4a021c3d
Binary files /dev/null and b/flow/designs/asap7/wns.png differ
diff --git a/flow/designs/gf12/README.md b/flow/designs/gf12/README.md
new file mode 100644
index 0000000000..3ad109e449
--- /dev/null
+++ b/flow/designs/gf12/README.md
@@ -0,0 +1,29 @@
+# gf12 designs
+
+<!-- BEGIN WNS (generated by flow/util/plot_wns.py) -->
+## WNS
+
+Worst setup slack per design at three flow stages — clock-tree synthesis (`cts`), global route (`globalroute`) and `finish` — read from each design's `rules-base.json`. Negative means setup timing is not met. Values are in this PDK's native timing unit (ps for `asap7`, ns for most others), so they are comparable within this PDK but not across PDKs.
+
+The bar is the `finish` slack; the markers show the `cts` and `globalroute` slack for the same design, so stage-to-stage movement is visible.
+
+![WNS by design — gf12](wns.png)
+
+| design | cts | globalroute | finish |
+| --- | ---: | ---: | ---: |
+| bp_single | -632 | -470 | -227 |
+| bp_quad | -375 | -223 | -225 |
+| ariane | -257 | -214 | -212 |
+| coyote | -200 | -200 | -200 |
+| bp_dual | -758 | -100 | -165 |
+| ca53 | -100 | -100 | -100 |
+| swerv_wrapper | -75 | -75 | -75 |
+| ibex | -51 | -51 | -51 |
+| tinyRocket | -40 | -40 | -40 |
+| jpeg | -96 | -53.2 | -25 |
+| aes | -21 | -21 | -21 |
+| gcd | -21.5 | -26.3 | -14 |
+| ariane133 | -20.2409 | 0 | 0 |
+
+_Generated by `flow/util/plot_wns.py` from `rules-base.json`; regenerate with `python3 flow/util/plot_wns.py`._
+<!-- END WNS -->
diff --git a/flow/designs/gf12/wns.png b/flow/designs/gf12/wns.png
new file mode 100644
index 0000000000..16f7673381
Binary files /dev/null and b/flow/designs/gf12/wns.png differ
diff --git a/flow/designs/gf180/README.md b/flow/designs/gf180/README.md
new file mode 100644
index 0000000000..2ac6540684
--- /dev/null
+++ b/flow/designs/gf180/README.md
@@ -0,0 +1,22 @@
+# gf180 designs
+
+<!-- BEGIN WNS (generated by flow/util/plot_wns.py) -->
+## WNS
+
+Worst setup slack per design at three flow stages — clock-tree synthesis (`cts`), global route (`globalroute`) and `finish` — read from each design's `rules-base.json`. Negative means setup timing is not met. Values are in this PDK's native timing unit (ps for `asap7`, ns for most others), so they are comparable within this PDK but not across PDKs.
+
+The bar is the `finish` slack; the markers show the `cts` and `globalroute` slack for the same design, so stage-to-stage movement is visible.
+
+![WNS by design — gf180](wns.png)
+
+| design | cts | globalroute | finish |
+| --- | ---: | ---: | ---: |
+| aes-hybrid | -0.967 | -1.1 | -1.1 |
+| aes | -0.789 | -0.876 | -0.918 |
+| riscv32i | -0.586 | -0.659 | -0.63 |
+| ibex | -0.5 | -0.5 | -0.539 |
+| jpeg | -0.375 | -0.375 | -0.375 |
+| uart-blocks | -0.3 | -0.3 | -0.3 |
+
+_Generated by `flow/util/plot_wns.py` from `rules-base.json`; regenerate with `python3 flow/util/plot_wns.py`._
+<!-- END WNS -->
diff --git a/flow/designs/gf180/wns.png b/flow/designs/gf180/wns.png
new file mode 100644
index 0000000000..0d07611130
Binary files /dev/null and b/flow/designs/gf180/wns.png differ
diff --git a/flow/designs/gf55/README.md b/flow/designs/gf55/README.md
new file mode 100644
index 0000000000..06d75913a3
--- /dev/null
+++ b/flow/designs/gf55/README.md
@@ -0,0 +1,17 @@
+# gf55 designs
+
+<!-- BEGIN WNS (generated by flow/util/plot_wns.py) -->
+## WNS
+
+Worst setup slack per design at three flow stages — clock-tree synthesis (`cts`), global route (`globalroute`) and `finish` — read from each design's `rules-base.json`. Negative means setup timing is not met. Values are in this PDK's native timing unit (ps for `asap7`, ns for most others), so they are comparable within this PDK but not across PDKs.
+
+The bar is the `finish` slack; the markers show the `cts` and `globalroute` slack for the same design, so stage-to-stage movement is visible.
+
+![WNS by design — gf55](wns.png)
+
+| design | cts | globalroute | finish |
+| --- | ---: | ---: | ---: |
+| aes |  |  | 0 |
+
+_Generated by `flow/util/plot_wns.py` from `rules-base.json`; regenerate with `python3 flow/util/plot_wns.py`._
+<!-- END WNS -->
diff --git a/flow/designs/gf55/wns.png b/flow/designs/gf55/wns.png
new file mode 100644
index 0000000000..bfc9693570
Binary files /dev/null and b/flow/designs/gf55/wns.png differ
diff --git a/flow/designs/gt2n/README.md b/flow/designs/gt2n/README.md
new file mode 100644
index 0000000000..459624ac72
--- /dev/null
+++ b/flow/designs/gt2n/README.md
@@ -0,0 +1,19 @@
+# gt2n designs
+
+<!-- BEGIN WNS (generated by flow/util/plot_wns.py) -->
+## WNS
+
+Worst setup slack per design at three flow stages — clock-tree synthesis (`cts`), global route (`globalroute`) and `finish` — read from each design's `rules-base.json`. Negative means setup timing is not met. Values are in this PDK's native timing unit (ps for `asap7`, ns for most others), so they are comparable within this PDK but not across PDKs.
+
+The bar is the `finish` slack; the markers show the `cts` and `globalroute` slack for the same design, so stage-to-stage movement is visible.
+
+![WNS by design — gt2n](wns.png)
+
+| design | cts | globalroute | finish |
+| --- | ---: | ---: | ---: |
+| aes | -25 | -75.2 | -75.2 |
+| jpeg | -50 | -50 | -50 |
+| gcd | -25 | -25 | -25 |
+
+_Generated by `flow/util/plot_wns.py` from `rules-base.json`; regenerate with `python3 flow/util/plot_wns.py`._
+<!-- END WNS -->
diff --git a/flow/designs/gt2n/wns.png b/flow/designs/gt2n/wns.png
new file mode 100644
index 0000000000..c299f3d54b
Binary files /dev/null and b/flow/designs/gt2n/wns.png differ
diff --git a/flow/designs/ihp-sg13g2/README.md b/flow/designs/ihp-sg13g2/README.md
new file mode 100644
index 0000000000..3b20f3eb50
--- /dev/null
+++ b/flow/designs/ihp-sg13g2/README.md
@@ -0,0 +1,23 @@
+# ihp-sg13g2 designs
+
+<!-- BEGIN WNS (generated by flow/util/plot_wns.py) -->
+## WNS
+
+Worst setup slack per design at three flow stages — clock-tree synthesis (`cts`), global route (`globalroute`) and `finish` — read from each design's `rules-base.json`. Negative means setup timing is not met. Values are in this PDK's native timing unit (ps for `asap7`, ns for most others), so they are comparable within this PDK but not across PDKs.
+
+The bar is the `finish` slack; the markers show the `cts` and `globalroute` slack for the same design, so stage-to-stage movement is visible.
+
+![WNS by design — ihp-sg13g2](wns.png)
+
+| design | cts | globalroute | finish |
+| --- | ---: | ---: | ---: |
+| i2c-gpio-expander | -1 | -1 | -1 |
+| ibex | -0.4 | -0.4 | -0.4 |
+| jpeg | -0.4 | -0.4 | -0.4 |
+| riscv32i | -0.3 | -0.3 | -0.3 |
+| aes | -0.225 | -0.225 | -0.225 |
+| gcd | -0.13 | -0.13 | -0.13 |
+| spi | 0 | -0.0426 | -0.045 |
+
+_Generated by `flow/util/plot_wns.py` from `rules-base.json`; regenerate with `python3 flow/util/plot_wns.py`._
+<!-- END WNS -->
diff --git a/flow/designs/ihp-sg13g2/wns.png b/flow/designs/ihp-sg13g2/wns.png
new file mode 100644
index 0000000000..3553d9a981
Binary files /dev/null and b/flow/designs/ihp-sg13g2/wns.png differ
diff --git a/flow/designs/nangate45/README.md b/flow/designs/nangate45/README.md
new file mode 100644
index 0000000000..d1e3aa2f57
--- /dev/null
+++ b/flow/designs/nangate45/README.md
@@ -0,0 +1,32 @@
+# nangate45 designs
+
+<!-- BEGIN WNS (generated by flow/util/plot_wns.py) -->
+## WNS
+
+Worst setup slack per design at three flow stages — clock-tree synthesis (`cts`), global route (`globalroute`) and `finish` — read from each design's `rules-base.json`. Negative means setup timing is not met. Values are in this PDK's native timing unit (ps for `asap7`, ns for most others), so they are comparable within this PDK but not across PDKs.
+
+The bar is the `finish` slack; the markers show the `cts` and `globalroute` slack for the same design, so stage-to-stage movement is visible.
+
+![WNS by design — nangate45](wns.png)
+
+| design | cts | globalroute | finish |
+| --- | ---: | ---: | ---: |
+| bp_quad | -150 | -150 | -150 |
+| black_parrot | -3.31 | -3.45 | -3.26 |
+| mempool_group | -2.31 | -2.31 | -2.31 |
+| swerv | -0.671 | -0.719 | -0.677 |
+| ariane133 | -0.579 | -0.569 | -0.595 |
+| swerv_wrapper | -0.344 | -0.357 | -0.35 |
+| dynamic_node | -0.362 | -0.38 | -0.344 |
+| ariane136 | -0.3 | -0.3 | -0.318 |
+| bp_be_top | -0.331 | -0.315 | -0.318 |
+| bp_multi_top | -0.24 | -0.24 | -0.24 |
+| jpeg | -0.147 | -0.165 | -0.164 |
+| tinyRocket | -0.14 | -0.168 | -0.154 |
+| bp_fe_top | -0.09 | -0.09 | -0.131 |
+| ibex | -0.11 | -0.127 | -0.11 |
+| aes | -0.041 | -0.0692 | -0.0667 |
+| gcd | -0.0529 | -0.0657 | -0.0559 |
+
+_Generated by `flow/util/plot_wns.py` from `rules-base.json`; regenerate with `python3 flow/util/plot_wns.py`._
+<!-- END WNS -->
diff --git a/flow/designs/nangate45/wns.png b/flow/designs/nangate45/wns.png
new file mode 100644
index 0000000000..8820d577f9
Binary files /dev/null and b/flow/designs/nangate45/wns.png differ
diff --git a/flow/designs/sky130hd/README.md b/flow/designs/sky130hd/README.md
new file mode 100644
index 0000000000..5825fee559
--- /dev/null
+++ b/flow/designs/sky130hd/README.md
@@ -0,0 +1,23 @@
+# sky130hd designs
+
+<!-- BEGIN WNS (generated by flow/util/plot_wns.py) -->
+## WNS
+
+Worst setup slack per design at three flow stages — clock-tree synthesis (`cts`), global route (`globalroute`) and `finish` — read from each design's `rules-base.json`. Negative means setup timing is not met. Values are in this PDK's native timing unit (ps for `asap7`, ns for most others), so they are comparable within this PDK but not across PDKs.
+
+The bar is the `finish` slack; the markers show the `cts` and `globalroute` slack for the same design, so stage-to-stage movement is visible.
+
+![WNS by design — sky130hd](wns.png)
+
+| design | cts | globalroute | finish |
+| --- | ---: | ---: | ---: |
+| microwatt | -2.71 | -2.8 | -2.73 |
+| gcd | -1.36 | -1.58 | -1.47 |
+| riscv32i | -1.41 | -1.55 | -1.37 |
+| jpeg | -0.716 | -0.877 | -0.654 |
+| ibex | -0.505 | -0.576 | -0.5 |
+| aes | -0.18 | -0.554 | -0.425 |
+| chameleon | -0.321 | -0.356 | -0.282 |
+
+_Generated by `flow/util/plot_wns.py` from `rules-base.json`; regenerate with `python3 flow/util/plot_wns.py`._
+<!-- END WNS -->
diff --git a/flow/designs/sky130hd/wns.png b/flow/designs/sky130hd/wns.png
new file mode 100644
index 0000000000..726d41557a
Binary files /dev/null and b/flow/designs/sky130hd/wns.png differ
diff --git a/flow/designs/sky130hs/README.md b/flow/designs/sky130hs/README.md
new file mode 100644
index 0000000000..4df31eea08
--- /dev/null
+++ b/flow/designs/sky130hs/README.md
@@ -0,0 +1,21 @@
+# sky130hs designs
+
+<!-- BEGIN WNS (generated by flow/util/plot_wns.py) -->
+## WNS
+
+Worst setup slack per design at three flow stages — clock-tree synthesis (`cts`), global route (`globalroute`) and `finish` — read from each design's `rules-base.json`. Negative means setup timing is not met. Values are in this PDK's native timing unit (ps for `asap7`, ns for most others), so they are comparable within this PDK but not across PDKs.
+
+The bar is the `finish` slack; the markers show the `cts` and `globalroute` slack for the same design, so stage-to-stage movement is visible.
+
+![WNS by design — sky130hs](wns.png)
+
+| design | cts | globalroute | finish |
+| --- | ---: | ---: | ---: |
+| gcd | -0.358 | -1 | -1 |
+| riscv32i | -0.303 | -0.695 | -0.571 |
+| ibex | -0.35 | -0.525 | -0.35 |
+| jpeg | -0.2 | -0.268 | -0.2 |
+| aes | -0.14 | -0.206 | -0.145 |
+
+_Generated by `flow/util/plot_wns.py` from `rules-base.json`; regenerate with `python3 flow/util/plot_wns.py`._
+<!-- END WNS -->
diff --git a/flow/designs/sky130hs/wns.png b/flow/designs/sky130hs/wns.png
new file mode 100644
index 0000000000..ad4963f3c3
Binary files /dev/null and b/flow/designs/sky130hs/wns.png differ
diff --git a/flow/designs/wns_accuracy.png b/flow/designs/wns_accuracy.png
new file mode 100644
index 0000000000..4c78bc48b6
Binary files /dev/null and b/flow/designs/wns_accuracy.png differ
diff --git a/flow/util/plot_wns.py b/flow/util/plot_wns.py
new file mode 100644
index 0000000000..9d613df325
--- /dev/null
+++ b/flow/util/plot_wns.py
@@ -0,0 +1,382 @@
+#!/usr/bin/env python3
+"""Plot worst setup slack (WNS) per design, per PDK, from committed rules-base.json.
+
+ORFS stores a golden-metrics baseline (rules-base.json) next to every design under
+flow/designs/<pdk>/<design>/. Each baseline records the worst setup slack reached at
+three flow stages:
+
+    cts__timing__setup__ws
+    globalroute__timing__setup__ws
+    finish__timing__setup__ws
+
+This script reads those values for every PDK that has them and regenerates, per PDK:
+
+    flow/designs/<pdk>/wns.png      -- horizontal bar chart of finish-stage WNS
+    flow/designs/<pdk>/README.md    -- a "## WNS" section (between generated markers)
+
+It also produces a cross-PDK view of how well the cts/globalroute estimates predict the
+final WNS (each design's per-stage estimate error, normalized by its clock period so the
+PDKs are comparable):
+
+    flow/designs/wns_accuracy.png   -- per-PDK strip plot of normalized estimate error
+    flow/designs/README.md          -- a "## WNS estimate accuracy across PDKs" section
+
+No OpenROAD/ORFS flow run is required -- the data is already in the tree, so the plots
+are deterministic and reproducible. Run from anywhere in the repo:
+
+    python3 flow/util/plot_wns.py            # all PDKs
+    python3 flow/util/plot_wns.py --pdk asap7  # one PDK
+
+Only matplotlib (pinned in flow/util/requirements_lock.txt) and the standard library are
+used.
+"""
+
+import argparse
+import glob
+import json
+import os
+import re
+import sys
+
+import matplotlib
+
+matplotlib.use("Agg")  # headless: write PNGs, never open a window
+import matplotlib.pyplot as plt  # noqa: E402
+
+# Stage key -> (short label, plot marker). Order is flow order.
+STAGES = [
+    ("cts__timing__setup__ws", "cts", "v"),
+    ("globalroute__timing__setup__ws", "globalroute", "^"),
+    ("finish__timing__setup__ws", "finish", None),  # finish is drawn as the bar
+]
+FINISH_KEY = "finish__timing__setup__ws"
+
+BEGIN = "<!-- BEGIN WNS (generated by flow/util/plot_wns.py) -->"
+END = "<!-- END WNS -->"
+
+ACC_BEGIN = "<!-- BEGIN WNS-ACCURACY (generated by flow/util/plot_wns.py) -->"
+ACC_END = "<!-- END WNS-ACCURACY -->"
+
+# Estimate stages whose accuracy (vs finish) we report, in flow order.
+EST_STAGES = ["cts", "globalroute"]
+EST_MARKERS = {"cts": "v", "globalroute": "^"}
+
+_PERIOD_RE = (
+    re.compile(r"set\s+clk_period\s+([0-9.]+)"),
+    re.compile(r"create_clock[^\n]*-period\s+([0-9.]+)"),
+)
+
+
+def clock_period(design_dir):
+    """Clock period for a design, parsed from its .sdc, or None if not found.
+
+    Handles the two idioms used across PDKs: `set clk_period <N>` and
+    `create_clock ... -period <N>`. Returns the first match (designs here are
+    single-clock); units are the PDK's native timing unit, same as the WNS values.
+    """
+    for sdc in sorted(glob.glob(os.path.join(design_dir, "*.sdc"))):
+        try:
+            text = open(sdc, errors="ignore").read()
+        except OSError:
+            continue
+        for rx in _PERIOD_RE:
+            m = rx.search(text)
+            if m:
+                return float(m.group(1))
+    return None
+
+
+def designs_dir():
+    """flow/designs, located relative to this script (flow/util/plot_wns.py)."""
+    here = os.path.dirname(os.path.abspath(__file__))
+    return os.path.normpath(os.path.join(here, "..", "designs"))
+
+
+def load_value(path, key):
+    """Return the float value for `key` in a rules-base.json, or None if absent."""
+    try:
+        with open(path) as f:
+            data = json.load(f)
+    except (OSError, ValueError):
+        return None
+    entry = data.get(key)
+    if isinstance(entry, dict) and isinstance(entry.get("value"), (int, float)):
+        return float(entry["value"])
+    return None
+
+
+def collect(pdk_dir):
+    """Return [(design, {stage_label: value_or_None}), ...] for designs with WNS data."""
+    rows = []
+    for design in sorted(os.listdir(pdk_dir)):
+        rules = os.path.join(pdk_dir, design, "rules-base.json")
+        if not os.path.isfile(rules):
+            continue
+        vals = {label: load_value(rules, key) for key, label, _ in STAGES}
+        if vals["finish"] is None:  # need at least the headline metric
+            continue
+        rows.append((design, vals))
+    # worst (most negative) finish slack first
+    rows.sort(key=lambda r: r[1]["finish"])
+    return rows
+
+
+def plot_pdk(pdk, rows, out_png):
+    names = [r[0] for r in rows]
+    finish = [r[1]["finish"] for r in rows]
+    y = list(range(len(names)))
+
+    fig_h = max(2.5, 0.42 * len(names) + 1.2)
+    fig, ax = plt.subplots(figsize=(9, fig_h))
+
+    colors = ["#c0392b" if v < 0 else "#27ae60" for v in finish]
+    ax.barh(y, finish, color=colors, height=0.6, zorder=2,
+            label="finish (bar)")
+
+    # overlay cts / globalroute as markers to show stage-to-stage movement
+    for key, label, marker in STAGES:
+        if marker is None:
+            continue
+        xs = [r[1][label] for r in rows]
+        yy = [yi for yi, x in zip(y, xs) if x is not None]
+        xx = [x for x in xs if x is not None]
+        if xx:
+            ax.scatter(xx, yy, marker=marker, s=40, zorder=3,
+                       edgecolors="black", linewidths=0.4, label=label)
+
+    ax.axvline(0, color="black", linewidth=0.8, zorder=1)
+    ax.set_yticks(y)
+    ax.set_yticklabels(names)
+    ax.invert_yaxis()  # worst slack on top
+    ax.set_xlabel("worst setup slack (PDK native timing unit; negative = not met)")
+    ax.set_title(f"WNS by design — {pdk}")
+    ax.grid(axis="x", linestyle=":", alpha=0.5, zorder=0)
+    ax.legend(loc="lower right", fontsize=8, framealpha=0.9)
+    fig.savefig(out_png, dpi=150, bbox_inches="tight")
+    plt.close(fig)
+
+
+def fmt(v):
+    return "" if v is None else f"{v:g}"
+
+
+def wns_section(pdk, rows):
+    lines = [
+        BEGIN,
+        "## WNS",
+        "",
+        "Worst setup slack per design at three flow stages — clock-tree synthesis "
+        "(`cts`), global route (`globalroute`) and `finish` — read from each "
+        "design's `rules-base.json`. Negative means setup timing is not met. Values are "
+        "in this PDK's native timing unit (ps for `asap7`, ns for most others), so they "
+        "are comparable within this PDK but not across PDKs.",
+        "",
+        "The bar is the `finish` slack; the markers show the `cts` and `globalroute` "
+        "slack for the same design, so stage-to-stage movement is visible.",
+        "",
+        f"![WNS by design — {pdk}](wns.png)",
+        "",
+        "| design | cts | globalroute | finish |",
+        "| --- | ---: | ---: | ---: |",
+    ]
+    for design, vals in rows:
+        lines.append(
+            f"| {design} | {fmt(vals['cts'])} | "
+            f"{fmt(vals['globalroute'])} | {fmt(vals['finish'])} |"
+        )
+    lines += [
+        "",
+        "_Generated by `flow/util/plot_wns.py` from `rules-base.json`; regenerate with "
+        "`python3 flow/util/plot_wns.py`._",
+        END,
+    ]
+    return "\n".join(lines) + "\n"
+
+
+def splice_readme(path, section, begin, end, title):
+    """Create README, or replace the marked section in place (preserving prose)."""
+    if os.path.isfile(path):
+        with open(path) as f:
+            text = f.read()
+        if begin in text and end in text:
+            pre = text[: text.index(begin)]
+            post = text[text.index(end) + len(end):]
+            new = pre + section + post.lstrip("\n")
+        else:
+            new = text.rstrip("\n") + "\n\n" + section
+    else:
+        new = (f"# {title}\n\n" if title else "") + section
+    with open(path, "w") as f:
+        f.write(new)
+
+
+# --- cross-PDK estimate accuracy --------------------------------------------
+
+def collect_accuracy(pdk_dir):
+    """Per-design normalized estimate error vs finish, in % of the clock period.
+
+    Returns [(design, {stage: err_pct}), ...] for designs that have a clock period and
+    finish + estimate-stage WNS. err_pct = 100 * (stage_ws - finish_ws) / period;
+    positive means the stage was *optimistic* (reported more slack than the final result).
+    Normalizing by clock period makes the error comparable across PDKs with different units.
+    """
+    out = []
+    for design in sorted(os.listdir(pdk_dir)):
+        ddir = os.path.join(pdk_dir, design)
+        rules = os.path.join(ddir, "rules-base.json")
+        if not os.path.isfile(rules):
+            continue
+        period = clock_period(ddir)
+        fin = load_value(rules, FINISH_KEY)
+        if not period or fin is None:
+            continue
+        errs = {}
+        for stage in EST_STAGES:
+            v = load_value(rules, f"{stage}__timing__setup__ws")
+            if v is not None:
+                errs[stage] = 100.0 * (v - fin) / period
+        if errs:
+            out.append((design, errs))
+    return out
+
+
+def _stats(rows, stage):
+    vals = [e[stage] for _, e in rows if stage in e]
+    if not vals:
+        return None
+    mae = sum(abs(v) for v in vals) / len(vals)
+    bias = sum(vals) / len(vals)
+    return len(vals), mae, bias, max(vals, key=abs)
+
+
+def plot_accuracy(acc, out_png):
+    """Strip plot: per-PDK distribution of cts/globalroute estimate error vs finish."""
+    pdks = sorted(acc)
+    colors = {"cts": "#2980b9", "globalroute": "#e67e22"}
+    off = {"cts": -0.18, "globalroute": 0.18}
+
+    fig, ax = plt.subplots(figsize=(max(8, 1.3 * len(pdks) + 2), 5.5))
+    for i, pdk in enumerate(pdks):
+        rows = acc[pdk]
+        for stage in EST_STAGES:
+            pts = [e[stage] for _, e in rows if stage in e]
+            k = len(pts)
+            xs = [
+                i + off[stage] + (0 if k < 2 else (j / (k - 1) - 0.5) * 0.26)
+                for j in range(k)
+            ]
+            ax.scatter(xs, pts, s=28, color=colors[stage], alpha=0.75,
+                       edgecolors="black", linewidths=0.3, zorder=3,
+                       label=stage if i == 0 else None)
+            if pts:  # mean tick
+                m = sum(pts) / k
+                ax.plot([i + off[stage] - 0.12, i + off[stage] + 0.12], [m, m],
+                        color=colors[stage], linewidth=2.5, zorder=4)
+
+    ax.axhline(0, color="black", linewidth=0.9, zorder=1)
+    ax.set_xticks(range(len(pdks)))
+    ax.set_xticklabels([f"{p}\n(n={len(acc[p])})" for p in pdks])
+    ax.set_ylabel("estimate − final WNS  (% of clock period)\n+ optimistic / − pessimistic")
+    ax.set_title("WNS estimate accuracy by stage, across PDKs")
+    ax.grid(axis="y", linestyle=":", alpha=0.5, zorder=0)
+    ax.legend(title="estimate stage", loc="upper left", fontsize=9)
+    fig.savefig(out_png, dpi=150, bbox_inches="tight")
+    plt.close(fig)
+
+
+def accuracy_section(acc):
+    lines = [
+        ACC_BEGIN,
+        "## WNS estimate accuracy across PDKs",
+        "",
+        "How closely the earlier-stage worst-slack estimates (`cts`, `globalroute`) match "
+        "the final (`finish`) WNS, per design, normalized by that design's clock period so "
+        "PDKs with different timing units are comparable. Error is "
+        "`(stage − finish) / clock_period`; **positive = optimistic** (the stage reported "
+        "more slack than the design actually closes with), negative = pessimistic. Clock "
+        "period is parsed from each design's `.sdc`; designs whose period could not be "
+        "parsed are omitted.",
+        "",
+        "![WNS estimate accuracy by stage, across PDKs](wns_accuracy.png)",
+        "",
+        "Mean absolute error (MAE) and mean signed error (bias), in % of clock period:",
+        "",
+        "| PDK | designs | cts MAE | cts bias | grt MAE | grt bias | worst (design) |",
+        "| --- | ---: | ---: | ---: | ---: | ---: | --- |",
+    ]
+    for pdk in sorted(acc):
+        rows = acc[pdk]
+        cs, gs = _stats(rows, "cts"), _stats(rows, "globalroute")
+        # worst single |error| over both stages, for context
+        worst = max(
+            ((abs(e[s]), e[s], d, s) for d, e in rows for s in e),
+            default=(0, 0, "-", ""),
+        )
+        c = f"{cs[1]:.1f}% | {cs[2]:+.1f}%" if cs else " | "
+        g = f"{gs[1]:.1f}% | {gs[2]:+.1f}%" if gs else " | "
+        lines.append(
+            f"| {pdk} | {len(rows)} | {c} | {g} | "
+            f"{worst[1]:+.1f}% ({worst[2]} {worst[3]}) |"
+        )
+    lines += [
+        "",
+        "_Generated by `flow/util/plot_wns.py`; regenerate with "
+        "`python3 flow/util/plot_wns.py`._",
+        ACC_END,
+    ]
+    return "\n".join(lines) + "\n"
+
+
+def main():
+    ap = argparse.ArgumentParser(description=__doc__)
+    ap.add_argument("--pdk", help="only process this PDK (default: all)")
+    args = ap.parse_args()
+
+    base = designs_dir()
+    if not os.path.isdir(base):
+        sys.exit(f"flow/designs not found at {base}")
+
+    pdks = sorted(
+        d for d in os.listdir(base)
+        if os.path.isdir(os.path.join(base, d))
+    )
+    if args.pdk:
+        pdks = [p for p in pdks if p == args.pdk]
+        if not pdks:
+            sys.exit(f"no such PDK directory: {args.pdk}")
+
+    processed = 0
+    for pdk in pdks:
+        pdk_dir = os.path.join(base, pdk)
+        rows = collect(pdk_dir)
+        if not rows:
+            continue
+        plot_pdk(pdk, rows, os.path.join(pdk_dir, "wns.png"))
+        splice_readme(os.path.join(pdk_dir, "README.md"), wns_section(pdk, rows),
+                      BEGIN, END, f"{pdk} designs")
+        miss = sum(1 for _, v in rows if v["finish"] < 0)
+        print(f"{pdk}: {len(rows)} designs ({miss} with negative finish WNS)")
+        processed += 1
+
+    if not processed:
+        sys.exit("no PDKs with rules-base.json WNS data found")
+
+    # Cross-PDK estimate-accuracy view (only meaningful over all PDKs).
+    if not args.pdk:
+        acc = {}
+        for pdk in pdks:
+            rows = collect_accuracy(os.path.join(base, pdk))
+            if rows:
+                acc[pdk] = rows
+        if acc:
+            plot_accuracy(acc, os.path.join(base, "wns_accuracy.png"))
+            splice_readme(os.path.join(base, "README.md"), accuracy_section(acc),
+                          ACC_BEGIN, ACC_END, "ORFS designs")
+            total = sum(len(v) for v in acc.values())
+            print(f"accuracy: {total} designs across {len(acc)} PDKs (clock period found)")
+
+    print(f"done: {processed} PDK(s)")
+
+
+if __name__ == "__main__":
+    main()