social-listening

Implements a Model Context Protocol server that standardizes connections to multiple social media platforms (Twitter/X, Instagram, TikTok, LinkedIn, etc.) through a unified MCP interface. Uses MCP's resource and tool abstractions to expose platform-specific APIs as composable, context-aware tools that Claude and other MCP-compatible clients can invoke. The server handles authentication token management, rate-limit coordination, and platform-specific payload normalization into a common schema.

Exposes search capabilities across social platforms with support for keyword queries, hashtag matching, account/user filtering, and temporal constraints. Implements query translation to platform-specific search syntax (Twitter's PowerTrack, Instagram's hashtag API, etc.) and aggregates results with consistent ranking/sorting. Handles pagination and result deduplication when querying multiple platforms simultaneously.

Analyzes posts retrieved from social platforms to extract sentiment (positive/negative/neutral), engagement metrics (likes, shares, comments, reach), and structured metadata (author influence, post type, media presence). Implements NLP-based sentiment classification (may use rule-based scoring, ML models, or platform-native sentiment APIs) and aggregates engagement data with time-series tracking. Extracts hashtags, mentions, URLs, and media references for downstream analysis.

Identifies emerging topics and trends from aggregated social media data by clustering posts with similar keywords, hashtags, or semantic content. Implements topic modeling (LDA, clustering algorithms) or keyword frequency analysis to surface trending discussions. Tracks trend velocity (growth rate), peak timing, and geographic/demographic distribution. Provides time-series data showing trend emergence and decay.

Retrieves and analyzes social media account profiles to extract influence metrics (follower count, engagement rate, audience demographics, posting frequency, content categories). Implements authority scoring based on follower growth, historical engagement, and network position. Provides audience composition data (age, location, interests) where available from platform APIs. Tracks account activity and content patterns over time.

Monitors mentions of competitor brands and products alongside your own brand, enabling comparative sentiment and engagement analysis. Implements mention deduplication (e.g., 'brand' vs 'Brand' vs 'BRAND'), competitor identification from a configurable list, and side-by-side metric comparison. Tracks competitive share of voice (percentage of mentions relative to total category mentions) and identifies posts mentioning multiple competitors simultaneously.

Monitors social media streams for specific events (mentions of brand/keywords, sentiment spikes, engagement thresholds, influencer activity) and triggers alerts when conditions are met. Implements configurable rules (e.g., 'alert if negative sentiment exceeds 30% in last hour', 'alert if post reaches 1000 likes in 10 minutes'). Supports multiple notification channels (webhook, email, Slack, etc.) and alert deduplication to prevent notification spam.

Analyzes historical social media performance data (engagement metrics, sentiment, reach) to recommend content types, posting times, hashtags, and messaging strategies. Implements pattern recognition to identify correlations between content attributes (post length, media type, hashtags, posting time) and engagement outcomes. Provides optimization suggestions (e.g., 'posts with 3-5 hashtags get 40% more engagement', 'video content performs 2x better than text-only on this audience').