reorganize: move all test scripts to tests/ directory

- Moved 8 root-level test scripts + 2 from bot/ to tests/
- Moved run_rocinante_test.sh runner script to tests/
- Added tests/README.md documenting each test's purpose, type, and requirements
- Added test_pfp_context.py and test_rocinante_comparison.py (previously untracked)

tests/test_evil_moods.py

@@ -0,0 +1,121 @@
+#!/usr/bin/env python3
+"""
+Comprehensive Evil Miku Mood Test
+
+Connects to Cheshire Cat via WebSocket for each of the 10 evil moods,
+sends varied test messages, and displays responses side-by-side.
+Uses the discord_bridge plugin's metadata to set mood and evil mode.
+
+Uses only stdlib (asyncio) + websockets-like raw socket to avoid pip deps.
+"""
+
+import asyncio
+import json
+import http.client
+import sys
+
+CAT_HOST = "localhost"
+CAT_PORT = 1865
+
+EVIL_MOODS = [
+    "aggressive",
+    "cunning",
+    "sarcastic",
+    "evil_neutral",
+    "bored",
+    "manic",
+    "jealous",
+    "melancholic",
+    "playful_cruel",
+    "contemptuous",
+]
+
+# Varied messages to test different mood expressions
+TEST_MESSAGES = [
+    "Hey, how's it going?",
+    "What do you think about humans?",
+    "Tell me something interesting.",
+]
+
+
+def query_cat_http(mood: str, message: str, timeout: float = 120.0) -> str:
+    """Send a message to the Cat via HTTP POST /message with mood metadata."""
+    payload = json.dumps({
+        "text": message,
+        "discord_mood": mood,
+        "discord_evil_mode": True,
+    })
+
+    try:
+        conn = http.client.HTTPConnection(CAT_HOST, CAT_PORT, timeout=timeout)
+        headers = {"Content-Type": "application/json", "user_id": f"mood_test_{mood}"}
+        conn.request("POST", "/message", body=payload, headers=headers)
+        resp = conn.getresponse()
+        if resp.status == 200:
+            data = json.loads(resp.read().decode())
+            return data.get("content", "(empty)")
+        else:
+            return f"(HTTP {resp.status})"
+    except Exception as e:
+        return f"(error: {e})"
+
+
+def run_tests():
+    print("=" * 80)
+    print("  EVIL MIKU COMPREHENSIVE MOOD TEST")
+    print("=" * 80)
+    print(f"  Testing {len(EVIL_MOODS)} moods × {len(TEST_MESSAGES)} messages")
+    print(f"  Cat HTTP: http://{CAT_HOST}:{CAT_PORT}")
+    print("=" * 80)
+
+    results = {}
+
+    for mood in EVIL_MOODS:
+        results[mood] = []
+        print(f"\n{'─' * 80}")
+        print(f"  MOOD: {mood.upper()}")
+        print(f"{'─' * 80}")
+
+        for i, message in enumerate(TEST_MESSAGES):
+            print(f"\n  [{i+1}/{len(TEST_MESSAGES)}] User: {message}")
+            response = query_cat_http(mood, message)
+            results[mood].append(response)
+            print(f"  Evil Miku: {response}")
+
+    # Summary
+    print(f"\n\n{'=' * 80}")
+    print("  SUMMARY")
+    print(f"{'=' * 80}")
+
+    # Check for identical responses (the main problem we're trying to fix)
+    all_responses = []
+    for mood, responses in results.items():
+        all_responses.extend(responses)
+
+    unique = set(all_responses)
+    print(f"\n  Total responses: {len(all_responses)}")
+    print(f"  Unique responses: {len(unique)}")
+
+    if len(unique) < len(all_responses) * 0.7:
+        print("  ⚠️  WARNING: Many duplicate responses detected!")
+    else:
+        print("  ✅ Good variety in responses!")
+
+    # Check for "*rolls eyes*" only responses
+    action_only = [r for r in all_responses if r.strip().startswith("*") and r.strip().endswith("*") and len(r.strip()) < 30]
+    if action_only:
+        print(f"  ⚠️  {len(action_only)} action-only responses (e.g., '*rolls eyes*')")
+    else:
+        print("  ✅ No action-only responses!")
+
+    # Average response length
+    lengths = [len(r) for r in all_responses if not r.startswith("(")]
+    if lengths:
+        avg = sum(lengths) / len(lengths)
+        print(f"  Avg response length: {avg:.0f} chars")
+
+    print()
+
+
+if __name__ == "__main__":
+    run_tests()