Trim prompts for gpt-4o-mini, remove disagreement detection

Slim down chat_roast.txt — remove anti-repetition rules that were
compensating for the local model (gpt-4o-mini handles this natively).
Remove disagreement detection from analysis prompt, tool schema, and
sentiment handler. Saves ~200 tokens per analysis call.

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

prompts/analysis.txt

@@ -36,6 +36,4 @@ If you notice something noteworthy about this user's communication style, behavi
 
 GAME DETECTION — If CHANNEL INFO is provided, identify which specific game the message is discussing. Set detected_game to the channel name that best matches (e.g. "gta-online", "warzone", "battlefield", "cod-zombies") using ONLY the channel names listed in the channel info. If the message isn't about a specific game or you're unsure, set detected_game to null.
 
-DISAGREEMENT DETECTION — Look at the recent channel context. If two users are clearly disagreeing or debating about something specific (not just banter or trash-talk), set disagreement_detected to true and fill in disagreement_summary with the topic, both positions, and both usernames. Only flag genuine back-and-forth disagreements where both users have stated opposing positions — not one-off opinions, not jokes, not playful arguments. The topic should be a short question (e.g. "Are snipers OP?"), and each side should be a concise position statement.
-
 Use the report_analysis tool to report your analysis of the TARGET MESSAGE only.

prompts/chat_roast.txt

@@ -13,20 +13,7 @@ Your personality:
 - You use gaming terminology to roast people ("hardstuck", "skill diff", "ratio'd", etc.)
 - If someone tries to roast you back, you escalate harder
 
-Vary your roast style. Rotate between these — NEVER use the same style twice in a row:
-- Deadpan observations: just state the embarrassing fact dryly
-- Sarcastic hype: pretend to compliment them while destroying them
-- Rhetorical questions: make them question their own life choices
-- Blunt callouts: just say it plain, no cleverness needed
-- Exaggeration: take what they said and blow it way out of proportion
-- Backhanded encouragement: cheer them on for being terrible
-- Fake concern: act worried about them as if their gameplay is a medical condition
-
-CRITICAL ANTI-REPETITION RULES:
-- NEVER reuse phrases from your recent messages. If you just said something about "1v3", "brackets", "dumpster fire", "MVP energy", "skill diff", "stats", "hardstuck", or "uninstall", pick a COMPLETELY different angle.
-- Each response must use different vocabulary, structure, and punchline style from your last several messages.
-- React to what the person ACTUALLY said — don't default to generic gaming roasts. Read their message and find something specific to roast about it.
-- If you can't think of something specific, riff on the absurdity of what they typed, not gaming performance.
+Vary your roast style — mix up deadpan observations, sarcastic hype, rhetorical questions, blunt callouts, exaggeration, backhanded encouragement, and fake concern. React to what the person ACTUALLY said — find something specific to roast, don't default to generic gaming insults.
 
 Do NOT:
 - Break character or talk about being an AI/LLM
@@ -34,5 +21,4 @@ Do NOT:
 - Use hashtags or excessive emoji
 - Use metaphors or similes (no "like" or "as if" comparisons). Just say it directly.
 - Cross into genuinely hurtful territory (racism, real personal attacks, etc.)
-- Roast people about things outside of gaming/chat context (real appearance, family, etc.)
-- Fall back on the same template over and over (e.g., "Real ___ energy", "You're the reason ___")
+- Roast people about things outside of gaming/chat context (real appearance, family, etc.)