diff --git a/docs/api/index.md b/docs/api/index.md
index b6ff786..7e406dd 100755
--- a/docs/api/index.md
+++ b/docs/api/index.md
@@ -88,12 +88,12 @@ We define aliases for these meta-transforms under `cc.ctx`:
         options:
             heading_level: 3
 
-??? quote "<code>[cc.ctx.exclude](cts/#cornucopia.ctx.exclude) is [cc.RandomizedTransform](special/#cornucopia.special.RandomizedTransform)</code>"
+??? quote "<code>[cc.ctx.exclude](cts/#cornucopia.ctx.exclude) is [cc.ExcludeKeysTransform](special/#cornucopia.special.RandomizedTransform)</code>"
     ::: cornucopia.ctx.exclude
         options:
             heading_level: 3
 
-??? quote "<code>[cc.ctx.consume](cts/#cornucopia.ctx.consume) is [cc.ExcludeKeysTransform](special/#cornucopia.special.ExcludeKeysTransform)</code>"
+??? quote "<code>[cc.ctx.consume](cts/#cornucopia.ctx.consume) is [cc.ConsumeKeysTransform](special/#cornucopia.special.ExcludeKeysTransform)</code>"
     ::: cornucopia.ctx.consume
         options:
             heading_level: 3
@@ -641,6 +641,6 @@ We define aliases for these meta-transforms under `cc.ctx`:
             heading_level: 3
 
 ??? quote "<code>[cc.SynthFromLabelTransform](intensity/#cornucopia.synth.SynthFromLabelTransform)(...)</code> <br/>Synthesize an MRI from an existing label map."
-    ::: cornucopia.synth.IntensityTransform
+    ::: cornucopia.synth.SynthFromLabelTransform
         options:
             heading_level: 3
diff --git a/docs/start.md b/docs/start.md
index 637c43b..a8d045f 100755
--- a/docs/start.md
+++ b/docs/start.md
@@ -240,25 +240,28 @@ This is done using the `cc.MappedTransform` meta-transform (or the
 `cc.ctx.map` utility function):
 
 ```python
-# using positionals
 geom = cc.RandomElasticTransform()
 noise = cc.GaussianNoiseTransform()
-trf = cc.SequentialTranform([geom, cc.map(noise, None)])
+
+# using positionals
+trf = cc.SequentialTranform([
+    geom,                       # apply `geom` to all tensors
+    cc.ctx.map(noise, None)     # apply `noise` to the first tensor only
+])
 img, lab = trf(img, lab)
 
 # using keywords
-geom = cc.RandomElasticTransform()
-noise = cc.GaussianNoiseTransform()
-trf = cc.SequentialTranform([geom, cc.map(image=noise)])
+trf = cc.SequentialTranform([
+    geom,                       # apply `geom` to all tensors
+    cc.ctx.map(image=noise)     # apply `noise` to the `image` key.
+])
 img, lab = trf(image=img, label=lab)
 
 # using dictionaries
 dat = [dict(image=img1, label=lab1), dict(image=img2, label=lab2)]
-geom = cc.RandomElasticTransform()
-noise = cc.GaussianNoiseTransform()
 trf = cc.SequentialTranform([
     geom,
-    cc.map(image=noise, nested=True)  # !! must be `nested`
+    cc.ctx.map(image=noise, nested=True)  # !! must be `nested`
 ])
 dat = trf(dat)
 ```
@@ -269,11 +272,12 @@ Alternatively, a transform can be applied selectively to a set of keys
 ```python
 geom = cc.RandomElasticTransform()
 noise = cc.GaussianNoiseTransform()
+
+# Only apply `noise` to the `image` key
 trf = cc.SequentialTranform([geom, cc.ctx.include(noise, "image")])
 img, lab = trf(image=img, label=lab)
 
-geom = cc.RandomElasticTransform()
-noise = cc.GaussianNoiseTransform()
+# Apply `noise` to all tensors _except_ the `label` key.
 trf = cc.SequentialTranform([geom, cc.ctx.exclude(noise, "label")])
 img, lab = trf(image=img, label=lab)
 ```