fix: tell headless Claude its text output IS the report file

claude -p captures stdout text, not tool output. The previous prompt
let Claude produce the detailed report via skill invocation but then
write only a summary as its final text response. Now explicitly states
that every word of text output goes into the saved report file.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: ZviBaratz

scripts/review-prompt.md

@@ -1,6 +1,6 @@
 Run /ego-review on the extension at {{EXT_PATH}}.
 
-IMPORTANT: You MUST read the extension source files and produce the FULL ego-review report with ALL six sections. Do NOT summarize. Do NOT skip phases. Output the complete report in markdown.
+Your ENTIRE text output will be saved as the review report file. Do NOT summarize at the end — your full output IS the deliverable. Write the complete report inline as you go, with all six sections in full detail. Every word you output goes into the report.
 
 ## Phase 0 Override
 
@@ -14,21 +14,28 @@ Skip Phase 0 (ego-lint). It has already been run. Use these pre-computed results
 
 {{ANNOTATIONS_SECTION}}
 
-## Required Output Format
+## Required Output
 
-You MUST output ALL of these sections with full detail:
+Write the FULL ego-review report with ALL sections below. Do NOT abbreviate or skip any section. If a section has no findings, write "None" under it — do not omit it.
+
+Start your output with:
+```
+## EGO Review Report — {{NAME}}
+```
+
+Then include ALL of these sections:
 
 ### Section 1: Blocking Issues (Must Fix)
-List each blocking issue with file:line, description, why reviewers reject it, and fix code.
+Each issue: file:line, description, why reviewers reject it, fix code.
 
 ### Section 2: Justification Required
 Items acceptable IF properly documented.
 
 ### Section 3: Advisory Issues (May Cause Questions)
-Each with file:line, description, reviewer perspective, suggestion.
+Each: file:line, description, reviewer perspective, suggestion.
 
 ### Section 4: Automated Check Summary
-Build from the lint JSON above. Table with Category × Pass/Fail/Warn.
+Build from the lint JSON above. Table: Category × Pass/Fail/Warn.
 
 ### Section 5: AI Pattern Analysis
 Score, triggered items with file:line, assessment.