feat: channel-level conversation analysis with compact formatting

Switch from per-user message batching to per-channel conversation
analysis. The LLM now sees the full interleaved conversation with
relative timestamps, reply chains, and consecutive message collapsing
instead of isolated flat text per user.

Key changes:
- Fix gpt-5-nano temperature incompatibility (conditional temp param)
- Add mention-triggered scan: users @mention bot to analyze recent chat
- Refactor debounce buffer from (channel_id, user_id) to channel_id
- Replace per-message analyze_message() with analyze_conversation()
  returning per-user findings from a single LLM call
- Add CONVERSATION_TOOL schema with coherence, topic, and game fields
- Compact message format: relative timestamps, reply arrows (→),
  consecutive same-user message collapsing
- Separate mention scan tasks from debounce tasks
- Remove _store_context/_get_context (conversation block IS the context)
- Escalation timeout config: [30, 60, 120, 240] minutes

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

config.yaml

@@ -33,8 +33,13 @@ topic_drift:
   escalation_count: 3  # After this many reminds, DM the server owner
   reset_minutes: 60  # Reset off-topic count after this much on-topic behavior
 
+mention_scan:
+  enabled: true
+  scan_messages: 30       # Messages to scan per mention trigger
+  cooldown_seconds: 60    # Per-channel cooldown between scans
+
 timeouts:
-  escalation_minutes: [5, 15, 30, 60]  # Escalating timeout durations
+  escalation_minutes: [30, 60, 120, 240]  # Escalating timeout durations
   offense_reset_minutes: 120  # Reset offense counter after this much good behavior
   warning_cooldown_minutes: 5  # Don't warn same user more than once per this window