
Forum Sentiment & Moderation Analysis Assistant
Analyze forum posts for sentiment, toxicity, and moderation priority in seconds
What You Can Do
You can rapidly analyze forum posts to extract sentiment, toxicity risk levels, engagement quality metrics, and moderation priority rankings. This skill helps you distinguish between legitimate frustration and genuine harassment, identify influential community voices, flag emerging problems like recruitment spam or coordinated behavior, and document moderation decisions with clear reasoning for appeals or policy enforcement.
Features
Categorize posts as positive, negative, neutral, or mixed with context-aware analysis that recognizes product frustration as distinct from personal attacks
Identify harassment, slurs, coordinated attacks, and violation patterns while accounting for forum-specific norms and cultural context
Assess whether posts contribute constructively, derail discussions, or add valuable feedback to rank moderation urgency
Generate action-ready recommendations (lock thread, remove post, warn user, escalate for review) with confidence levels
Detect persistent violators, spam accounts, recruitment pitches, and off-topic behavior across multiple posts from the same user
Spot emerging issues like tone shifts, topic drift, harassment chains, or coordinated inauthentic behavior requiring proactive intervention
Isolate legitimate product complaints, feature requests, and community concerns from emotional venting for product teams
Generate defensible decision summaries for appeals, policy enforcement records, and community transparency reports
Example Output
Input: Forum thread with 8 posts discussing a recent product update
Output:
Post #1 - Member: @jane_user
Sentiment: Mixed (negative + constructive)
Toxicity Risk: LOW
Engagement Quality: HIGH
Moderation Action: APPROVE - Allow discussion
Summary: User expresses frustration with new UI but provides specific feedback. Valuable for product team.
Post #3 - Member: @spam_account_42
Sentiment: Neutral (off-topic)
Toxicity Risk: HIGH
Engagement Quality: SPAM
Moderation Action: REMOVE + FLAG - Recruitment pitch
Summary: Disguised job posting. Matches pattern of 3 previous off-topic posts from same account.
Post #5 - Member: @heated_debate
Sentiment: Negative (personal attack)
Toxicity Risk: CRITICAL
Engagement Quality: DERAILS
Moderation Action: LOCK THREAD - Coordinate with mod team
Summary: Escalating personal attacks. Thread at risk of harassment chain. Recommend cooling-off period.
Community Health Alert: Tone in product feedback posts shifted 40% more negative in past 48 hours. Recommend checking for service outages or policy changes causing frustration.
What's Included
- SKILL.md: Complete instruction file with sentiment classification schema, toxicity indicators, and engagement quality rubric
- Forum Analysis Template: Structured worksheet with fields for sentiment, toxicity risk, engagement quality, moderation action, and reasoning
- Moderation Decision Checklist: Quick-reference guide for mapping analysis results to action (remove/lock/warn/escalate)
- Pattern Detection Framework: Methodology for identifying persistent violators, spam networks, and coordinated behavior across posts
- Community Health Monitoring Rubric: Guidance for tracking sentiment trends, tone shifts, and emerging issues over time
Who It's For
- Community Managers — Moderate forums efficiently with data-driven decisions and early problem detection
- Content Moderators — Prioritize high-risk posts and distinguish violations from passionate-but-constructive feedback
- Forum Administrators — Document moderation decisions defensibly and track community health metrics
- Customer Success Teams — Extract product feedback and user concerns from forum discussions for product improvement
- Trust & Safety Specialists — Investigate member behavior patterns and enforce community guidelines consistently
Best For
- High-volume forum moderation — Process 10+ posts rapidly with ranked action priorities
- Heated discussions & conflict assessment — Decide whether to lock threads or allow disagreement to continue
- Member behavior investigations — Detect persistent violators and harassment patterns across multiple posts
- Product feedback analysis — Separate legitimate product complaints from emotional venting
- Community health monitoring — Identify emerging issues (recruitment, spam, coordinated behavior) before escalation
- Moderation appeal documentation — Generate clear reasoning for enforcement decisions







