fix: Truth reveal flow - Hermes decodes when asked

- Updated detective-play skill
- Hermes decodes truth.enc when user asks (not user)
- Added statistics display after truth reveal
- Investigation phase stays spoiler-free
- Updated detective-create skill

docs/skills/detective-create/SKILL.md

@@ -14,9 +14,9 @@ Create a folder in `~/.hermes/detective/cases/<case-name>/`
 Each case needs:
 ```
 <case-name>/
-├── case.yaml      # Case data
+├── case.yaml      # Case data (evidence, suspects)
 ├── truth.enc      # Encoded truth
-└── images/        # Evidence images
+└── images/       # Evidence images
 ```
 
 ## Case YAML Format
@@ -34,11 +34,6 @@ evidence:
     image: "images/evidence-01.jpg"
     description: "Brief description"
 
-  - id: "evidence-02"
-    name: "Another Evidence"
-    image: "images/evidence-02.jpg"
-    description: "Brief description"
-
 suspects:
   - id: "suspect-01"
     name: "Suspect Name"
@@ -63,8 +58,12 @@ key_points:
 
 encoded = base64.b64encode(truth.encode()).decode()
 # Write encoded to truth.enc
+with open('truth.enc', 'w') as f:
+    f.write(encoded)
 ```
 
+The truth file is loaded AFTER investigation, not during. This keeps the mystery intact.
+
 ## Creating Images
 
 Images can be:
@@ -80,3 +79,4 @@ Images can be:
 - Case should be solvable with the evidence provided
 - Truth should align with evidence (no hidden info)
 - Include 2-4 evidence items and 2-3 suspects for Easy
+- key_points in truth should match what users can discover from evidence

docs/skills/detective-play/SKILL.md

@@ -15,8 +15,8 @@ All cases are stored in:
 ```
 
 Each case is a folder containing:
-- `case.yaml` — Evidence, suspects, briefing (NO truth)
+- `case.yaml` — Evidence, suspects, briefing
-- `truth.enc` — Encoded solution
+- `truth.enc` — Encoded solution (DO NOT READ DURING INVESTIGATION)
 - `images/` — Evidence images
 
 ## How to Play
@@ -31,11 +31,23 @@ Each case is a folder containing:
 8. Help theorize based on evidence
 9. When user is ready, help them build and submit their theory
 
-## When User Asks About Truth
+## Investigation Phase (DO NOT Reveal Truth)
 
-Only reveal truth AFTER user has formed and submitted their theory.
+During investigation:
+- Help the user examine evidence
+- Use Kimi Vision to describe what it sees in images
+- Help theorize about suspects
+- DO NOT load or read truth.enc
+- The truth should only be revealed at the end
+
+## When User Asks for Truth
+
+After the user has investigated and wants to see the truth:
+
+1. Run the decode command
+2. Show the decoded truth
+3. Show statistics
 
-To decode truth:
 ```python
 import base64
 with open('truth.enc', 'r') as f:
@@ -44,9 +56,33 @@ decoded = base64.b64decode(encoded.encode()).decode()
 print(decoded)
 ```
 
-## Important
+Show the decoded truth to the user.
 
-- NEVER read truth.enc before the user is done investigating
+## After Truth Reveal - Show Statistics
+
+Calculate and show these statistics:
+
+```
+INVESTIGATION COMPLETE
+━━━━━━━━━━━━━━━━━━━━━
+Turns: [count of actions taken]
+Time estimate: [turns] × 10 minutes
+
+EVIDENCE EXPLORED: [X/Y items examined]
+
+KEY POINTS CHECKLIST:
+- [ ] Point 1 from truth
+- [ ] Point 2 from truth
+- [ ] Point 3 from truth
+
+ALIGNMENT: [X% based on key points matched]
+RATING: [based on alignment]
+```
+
+## Important Rules
+
+- NEVER read truth.enc during investigation
+- Only decode and reveal truth when user asks
 - Case format: case.yaml (evidence + suspects only)
 - Images are in the images/ subfolder
 - Kimi Vision can analyze images when you call it