diff --git a/skills/subagent-driven-development/SKILL.md b/skills/subagent-driven-development/SKILL.md
index 67257e18..c21dce50 100644
--- a/skills/subagent-driven-development/SKILL.md
+++ b/skills/subagent-driven-development/SKILL.md
@@ -282,6 +282,9 @@ Done!
 - Tell a reviewer what not to flag, or pre-rate a finding's severity in the
   dispatch prompt ("treat it as Minor at most") — the plan's example code is
   a starting point, not evidence that its weaknesses were chosen
+- Dispatch a task reviewer without pasting the diff into the prompt — run
+  `git diff BASE..HEAD` yourself first (`--stat` plus relevant hunks if it
+  exceeds a few hundred lines)
 - Move to next task while the review has open Critical/Important issues
 
 **If subagent asks questions:**
diff --git a/skills/subagent-driven-development/task-reviewer-prompt.md b/skills/subagent-driven-development/task-reviewer-prompt.md
index 12c6da51..c3ead9cb 100644
--- a/skills/subagent-driven-development/task-reviewer-prompt.md
+++ b/skills/subagent-driven-development/task-reviewer-prompt.md
@@ -28,7 +28,9 @@ Subagent (general-purpose):
     **Base:** [BASE_SHA]
     **Head:** [HEAD_SHA]
 
-    [DIFF]
+    [DIFF — REQUIRED: the controller pastes `git diff BASE..HEAD` output
+    here before dispatching. Dispatching with this placeholder unfilled is
+    a dispatch error.]
 
     Review from the diff above — do not re-run git commands or re-read the
     files it already shows. If no diff was provided, fetch it yourself:
@@ -148,9 +150,9 @@ Subagent (general-purpose):
 - `[DESCRIPTION]` — what the implementer reports they built
 - `[BASE_SHA]` — commit before this task
 - `[HEAD_SHA]` — current commit
-- `[DIFF]` — paste `git diff BASE..HEAD` output (use `--stat` plus the
-  relevant hunks if it exceeds a few hundred lines); a reviewer with the
-  diff in hand needs few or no tool calls
+- `[DIFF]` — REQUIRED: paste `git diff BASE..HEAD` output (use `--stat`
+  plus the relevant hunks if it exceeds a few hundred lines); a reviewer
+  with the diff in hand needs few or no tool calls
 
 **Reviewer returns:** Spec Compliance verdict (✅/❌/⚠️), Strengths, Issues
 (Critical/Important/Minor), Task quality verdict