Merge branch 'feature/v0.4.0-profile' into main

v0.4.0 profile improvements:
- 19+ title variety (Discord Whisperer no longer dominant)
- top-2 shills + top-2 haters per user (multiple brand pills)
- posting style as multiple pills (one per trait, flex-wrap)
- brand lover/hater as AWARD (Mimo hater, MiniMax lover format)
- awards badge legend now includes all new awards
- Mod + Developer + Contributor pills (hardcoded user lists)
- removed Vibe section (was always 'mood unavailable')

Author: Hermes

build_index.py

@@ -160,20 +160,69 @@ def _run(args):
 HATER_MIN_NEG_RATIO = 5.0     # Min negative:positive ratio to qualify as hater (was 1:3)
 USE_WORD_BOUNDARY = True       # Use word-boundary regex in brand keyword matching
 
-# ─── v0.3.0: Title assignment rules (ISSUE #2) ─────────────────────
-# Priority order (first match wins). Each entry is (label, emoji, condition_fn_string).
-# Conditions are checked in order top-to-bottom.
+# ─── v0.4.0: Title assignment rules (ISSUE #1) ─────────────────────
+def _peak_hour_bucket(s):
+    """Return time-of-day bucket name or None."""
+    hour_dist = s.get("hour_dist", {}) or {}
+    if not hour_dist:
+        return None
+    peak = max(hour_dist, key=hour_dist.get)
+    if peak < 6:
+        return "early_bird"
+    if peak < 12:
+        return "morning"
+    if peak < 14:
+        return "lunch"
+    if peak < 18:
+        return "afternoon"
+    return "night_owl"
+
 TITLE_RULES = [
-    # (priority_order, "title", "emoji", condition_description)
-    # Each checked in assign_title() in priority order
+    # (priority, name, emoji, check_fn(stats_dict, rank_idx))
+    # Lower number = more specific, checked first
+    (1, "Foundation", "\U0001F451",
+     lambda s, r: r is not None and r <= TITLE_FOUNDATION_RANK),
+    (2, "OG", "\U0001F3DB\uFE0F",
+     lambda s, r: ((s.get("first_seen") or "9999")[:10] < TITLE_OG_FIRST_SEEN
+                   and (s.get("total_chunks") or 0) >= TITLE_OG_MIN_CHUNKS)),
+    (3, "Veteran", "\U0001F396\uFE0F",
+     lambda s, r: ((s.get("first_seen") or "9999")[:10] < TITLE_VETERAN_FIRST_SEEN
+                   and (s.get("max_streak") or 0) >= TITLE_VETERAN_MIN_STREAK)),
+    (4, "Pioneer", "\U0001F48E",
+     lambda s, r: ((s.get("first_seen") or "9999")[:10] < TITLE_PIONEER_FIRST_SEEN)),
+    (5, "Globetrotter", "\U0001F30D",
+     lambda s, r: (len(s.get("channels", set()) or set()) >= TITLE_GLOBETROTTER_MIN_CHANNELS
+                   and (s.get("total_chunks") or 0) >= TITLE_GLOBETROTTER_MIN_CHUNKS)),
+    (6, "Trailblazer", "\U0001F680",
+     lambda s, r: (s.get("total_chunks") or 0) >= TITLE_TRAILBLAZER_MIN_CHUNKS),
+    (7, "Discord Whisperer", "\U0001F5E3\uFE0F",
+     lambda s, r: ((s.get("msg_per_chunk") or 0) > TITLE_WHISPERER_MIN_MSG_RATIO
+                   and (s.get("total_chunks") or 0) >= TITLE_WHISPERER_MIN_CHUNKS)),
+    (8, "Polymath", "\U0001F52C",
+     lambda s, r: len(s.get("channels", set()) or set()) >= TITLE_POLYMATH_MIN_CHANNELS),
+    (9, "Sage", "\U0001F9E0",
+     lambda s, r: ((s.get("msg_per_chunk") or 0) > TITLE_SAGE_MIN_MSG_PER_CHUNK
+                   and (s.get("total_chunks") or 0) >= TITLE_SAGE_MIN_CHUNKS)),
+    (10, "Night Owl", "\U0001F989",
+     lambda s, r: _peak_hour_bucket(s) == "night_owl"),
+    (11, "Early Bird", "\U0001F305",
+     lambda s, r: _peak_hour_bucket(s) == "early_bird"),
+    (12, "Morning Poster", "\u2615",
+     lambda s, r: _peak_hour_bucket(s) == "morning"),
+    (13, "Lunch Break", "\U0001F96A",
+     lambda s, r: _peak_hour_bucket(s) == "lunch"),
+    (14, "Afternoon Poster", "\U0001F31E",
+     lambda s, r: _peak_hour_bucket(s) == "afternoon"),
+    (15, "Lurker", "\U0001F575\uFE0F",
+     lambda s, r: ((s.get("total_chunks") or 0) >= TITLE_LURKER_MIN_CHUNKS
+                   and (s.get("msg_per_chunk") or 0) < TITLE_LURKER_MAX_MSG_RATIO)),
+    (16, "Linker", "\U0001F517",
+     lambda s, r: (s.get("link_rate") or 0) > TITLE_LINKER_MIN_RATE),
+    (17, "Code Wizard", "\U0001F9D9",
+     lambda s, r: (s.get("code_rate") or 0) > TITLE_CODE_WIZARD_MIN_RATE),
+    (18, "Builder", "\U0001F3D7\uFE0F",
+     lambda s, r: (s.get("total_chunks") or 0) >= TITLE_BUILDER_MIN_CHUNKS),
 ]
-# Title thresholds: checked by assign_title(). Rarest/most prestigious first.
-# Top 5 by rank -> "Pioneer" (uses rank_idx, handled during leaderboard building)
-# msg_per_chunk > 15 -> "Sage"
-# streaks >= 30 -> "Streak Master"
-# 100+ chunks -> "Trailblazer"
-# 4+ channels active -> "Polymath"
-# 50+ chunks -> "Builder"
 
 # ─── v0.3.0: GPU tier emojis (ISSUE #8, #17) ───────────────────────
 GPU_TIER_EMOJI = {
@@ -187,6 +236,39 @@ def _run(args):
     "integrated": "\U0001F4BB",
 }
 
+# ─── v0.4.0: Title rule thresholds (ISSUE #1) ────────────
+TITLE_FOUNDATION_RANK = 5
+TITLE_OG_FIRST_SEEN = "2025-01-01"
+TITLE_OG_MIN_CHUNKS = 50
+TITLE_PIONEER_FIRST_SEEN = "2025-01-01"
+TITLE_VETERAN_FIRST_SEEN = "2025-06-01"
+TITLE_VETERAN_MIN_STREAK = 30
+TITLE_TRAILBLAZER_MIN_CHUNKS = 100
+TITLE_BUILDER_MIN_CHUNKS = 50
+TITLE_POLYMATH_MIN_CHANNELS = 4
+TITLE_SAGE_MIN_MSG_PER_CHUNK = 15
+TITLE_SAGE_MIN_CHUNKS = 10
+TITLE_WHISPERER_MIN_MSG_RATIO = 20
+TITLE_WHISPERER_MIN_CHUNKS = 20
+TITLE_LURKER_MIN_CHUNKS = 50
+TITLE_LURKER_MAX_MSG_RATIO = 3
+TITLE_LINKER_MIN_RATE = 0.15
+TITLE_CODE_WIZARD_MIN_RATE = 0.3
+TITLE_EMOJI_MASTER_MIN_RATE = 0.4
+TITLE_QUESTIONER_MIN_RATE = 0.4
+TITLE_GLOBETROTTER_MIN_CHANNELS = 4
+TITLE_GLOBETROTTER_MIN_CHUNKS = 50
+
+# ─── v0.4.0: Brand lover/hater thresholds (ISSUE #13) ──────
+BRAND_LOVER_AWARD_MIN_TOTAL = 10
+BRAND_LOVER_AWARD_MIN_POS_RATIO = 3.0
+BRAND_HATER_AWARD_MIN_TOTAL = 10
+BRAND_HATER_AWARD_MIN_NEG_RATIO = 3.0
+
+# ─── v0.4.0: Known community roles (ISSUE #15) ─────────────
+MOD_USERS = {"teknium"}
+DEVELOPER_USERS = {"4rgo", "teknium"}
+
 # ─── T11: GPU power ranking list ────────────────────────────────────
 GPU_LIST = [
     (  1, "GB200 Grace+Blackwell",  "384GB (2x B200)",  "frontier", "#facc15"),
@@ -517,12 +599,14 @@ def assign_medal(rank, name):
     return ["ribbon", "Ribbon"]
 
 
-def assign_awards(user_stats):
+def assign_awards(user_stats, author_name=None):
     """Stable contribution awards from lifetime stats. Only go up.
 
     ISSUE #4: 8 new awards added (Linker, Convo-starter, Emoji Master,
     Questioner, Helper, Mentor, OG, Resurrected). Existing awards have
     priority (checked first). Cap at 2 awards per user.
+    ISSUE #13: Brand lover/hater awards added.
+    ISSUE #15: Mod/Developer/Contributor role pills added (separate field).
     """
     awards = []
     channels = len(user_stats.get("channels", set()) or set())
@@ -587,33 +671,31 @@ def assign_awards(user_stats):
         if max_gap_days > 120:
             awards.append(["resurrected", "Resurrected"])
 
+    # ISSUE #13: Brand lover/hater awards (checked after standard awards)
+    if len(awards) < 2:
+        _shills = user_stats.get("shill_brands", []) or []
+        for _s in _shills:
+            _bm = BRAND_META.get(_s["brand"])
+            if _bm and (_s.get("count") or 0) >= BRAND_LOVER_AWARD_MIN_TOTAL:
+                awards.append(["lover", f"{_bm[1]} lover"])
+                break
+
+    if len(awards) < 2:
+        _haters = user_stats.get("hater_brands", []) or []
+        for _h in _haters:
+            _bm = BRAND_META.get(_h["brand"])
+            if _bm and (_h.get("count") or 0) >= BRAND_HATER_AWARD_MIN_TOTAL:
+                awards.append(["hater", f"{_bm[1]} hater"])
+                break
+
     return awards[:2]
 
 
 def assign_title(s, rank_idx=None):
-    """Return a single per-user title string based on strongest contribution signal.
-
-    ISSUE #2: Priority-ordered checks. First match wins.
-    Titles: Pioneer, Sage, Streak Master, Trailblazer, Polymath, Builder.
-    """
-    chunks = s.get("total_chunks", 0) or 0
-    msgs = s.get("total_messages", 0) or 0
-    channels_n = len(s.get("channels", set()) or set())
-    msg_per_chunk = s.get("msg_per_chunk", 0) or 0
-    max_streak = s.get("max_streak", 0) or 0
-
-    if rank_idx is not None and rank_idx <= 5:
-        return "\U0001F48E Pioneer"
-    if msg_per_chunk > 15 and chunks > 10:
-        return "\U0001F9E0 Sage"
-    if max_streak >= 30:
-        return "\U0001F525 Streak Master"
-    if chunks >= 100:
-        return "\U0001F680 Trailblazer"
-    if channels_n >= 4:
-        return "\U0001F9EC Polymath"
-    if chunks >= 50:
-        return "\U0001F6E0\uFE0F Builder"
+    """Evaluate all TITLE_RULES and return the best match (highest priority)."""
+    for _priority, _name, _emoji, _check_fn in TITLE_RULES:
+        if _check_fn(s, rank_idx):
+            return f"{_emoji} {_name}"
     return None
 
 
@@ -706,6 +788,74 @@ def compute_style_heuristic(s):
     return None
 
 
+def compute_style_pills(s):
+    """Return a list of individual style pill strings (ISSUE #10).
+
+    Each pill is one trait: channel activity, time of day, message length,
+    linker, code wizard, emoji master, questioner.
+    """
+    if not s.get("total_chunks"):
+        return []
+    pills = []
+    chunks = s.get("total_chunks", 0)
+    msgs = s.get("total_messages", 0)
+    chans = sorted(s.get("channels", set()) or set())
+    chan_short = {"hermes-agent": "hermes", "community-projects-showcase": "projects",
+                  "plugins-skills-and-skins": "plugins", "developers": "devs"}
+    chans_disp = [chan_short.get(c, c) for c in chans]
+
+    # Channel activity
+    if chunks <= 3:
+        if chans_disp:
+            chan_str = f"in #{chans_disp[0]}" if len(chans_disp) == 1 else f"in #{' + #'.join(chans_disp)}"
+            pills.append(f"Brief contributor {chan_str}")
+        return pills
+    if len(chans_disp) == 1:
+        pills.append(f"active in #{chans_disp[0]}")
+    elif len(chans_disp) <= 3:
+        pills.append(f"active in #{' + #'.join(chans_disp)}")
+    else:
+        pills.append(f"polymath across all {len(chans_disp)} channels")
+
+    # Time of day
+    hour_dist = s.get("hour_dist", {}) or {}
+    if hour_dist:
+        peak_hour = max(hour_dist, key=hour_dist.get)
+        if peak_hour < 6:
+            pills.append("night owl (UTC)")
+        elif peak_hour < 12:
+            pills.append("morning poster (UTC)")
+        elif peak_hour < 18:
+            pills.append("afternoon poster (UTC)")
+        else:
+            pills.append("evening poster (UTC)")
+
+    # Message length
+    avg_msg_len = round(s.get("total_msg_len", 0) / max(msgs, 1))
+    if avg_msg_len > 350:
+        pills.append(f"long-form (avg {avg_msg_len} chars)")
+    elif avg_msg_len < 100:
+        pills.append(f"concise (avg {avg_msg_len} chars)")
+    else:
+        pills.append(f"mid-length (avg {avg_msg_len} chars)")
+
+    # Style traits
+    link_rate = s.get("link_rate", 0) or 0
+    if link_rate > 0.15:
+        pills.append("linker")
+    code_rate = s.get("code_rate", 0) or 0
+    if code_rate > 0.3:
+        pills.append("code wizard")
+    emoji_rate = s.get("emoji_rate", 0) or 0
+    if emoji_rate > 0.4:
+        pills.append("emoji master")
+    q_rate = s.get("question_rate", 0) or 0
+    if q_rate > 0.4:
+        pills.append("questioner")
+
+    return pills
+
+
 # ─── T10: Brand detection ─────────────────────────────────────────
 def _match_brand_kw(content, kw):
     """Match a keyword against content — supports \\b regex patterns."""
@@ -753,17 +903,17 @@ def assign_shill_hater(brand_sentiment):
             continue
         # All positive (no negative mentions ever)
         if pos > 0 and neg == 0 and pos >= BRAND_MENTION_THRESHOLD:
-            shill.append({"brand": brand, "count": pos, "ratio": 9999.99})
+            shill.append({"brand": brand, "count": pos, "ratio": 9999.99, "total": total})
         # All negative (no positive mentions ever)
         elif neg > 0 and pos == 0 and neg >= BRAND_MENTION_THRESHOLD:
-            hater.append({"brand": brand, "count": neg, "ratio": 0.0})
+            hater.append({"brand": brand, "count": neg, "ratio": 0.0, "total": total})
         # Mixed: check ratios
         elif pos + neg >= BRAND_MENTION_THRESHOLD:
             ratio = pos / max(neg, 1)
             if ratio >= SHILL_MIN_POS_RATIO:
-                shill.append({"brand": brand, "count": pos, "ratio": round(ratio, 2)})
+                shill.append({"brand": brand, "count": pos, "ratio": round(ratio, 2), "total": total})
             elif ratio <= (1.0 / HATER_MIN_NEG_RATIO):
-                hater.append({"brand": brand, "count": neg, "ratio": round(ratio, 2)})
+                hater.append({"brand": brand, "count": neg, "ratio": round(ratio, 2), "total": total})
     return shill, hater
 
 
@@ -1191,8 +1341,11 @@ def main():
     for rank_idx, (author, s) in enumerate(_ranked, start=1):
         tc = s["total_chunks"]
         medal = assign_medal(rank_idx, author)
-        awards = assign_awards(s)
+        awards = assign_awards(s, author)
         title = assign_title(s, rank_idx)
+        # Cap shill/hater at top-2 by total mention count (ISSUE #9)
+        _shills = sorted(s.get("shill_brands", []) or [], key=lambda x: x.get("total", 0), reverse=True)[:2]
+        _haters = sorted(s.get("hater_brands", []) or [], key=lambda x: x.get("total", 0), reverse=True)[:2]
         leaderboard.append({
             "name": author, "total_chunks": tc,
             "total_messages": s["total_messages"],
@@ -1206,10 +1359,13 @@ def main():
             "awards": awards,
             "title": title,
             "style_heuristic": compute_style_heuristic(s),
+            "style_pills": compute_style_pills(s),
             "style_summary": STYLE_CACHE.get(author, {}).get("summary"),
-            "shill": s.get("shill_brands", []),
-            "hater": s.get("hater_brands", []),
+            "shill": _shills,
+            "hater": _haters,
             "developed": s.get("developed_repos", []),
+            "mod": author in MOD_USERS,
+            "developer": author in DEVELOPER_USERS,
         })
 
     # Per-user mood context (T14a) — last 20 chunks sorted by recency