social-media-intelligence

Assess account authenticity, detect coordinated behavior, and track narrative spread across social platforms.

43|8|Updated Mar 25, 2026
One-click install
npx skills add https://github.com/buriedsignals/spotlight --skill social-media-intelligence-buriedsignals
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: social-media-intelligence
Source: https://github.com/buriedsignals/spotlight/tree/main/skills/social-media-intelligence
Command: npx skills add https://github.com/buriedsignals/spotlight --skill social-media-intelligence-buriedsignals

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Investigators and journalists struggle to determine whether social media accounts are authentic, whether multiple accounts are coordinating to amplify a narrative, and how a claim actually spread across platforms before the evidence disappears. ## Core Features & Use Cases - Account Authenticity Assessment: Systematic red-flag checklist (account age, follower ratios, posting volume, profile photos) with a scoring rule that ties three or more flags to explicit low-confidence findings. - Coordination Detection: Checklists for timing, content, account, and network patterns with thresholds that determine whether accounts should be treated as a single source. - Narrative Tracking: Four-step method to find a claim's origin, map its spread with structured JSON records, identify amplifiers, and measure propagation velocity. - Platform Collection Guidance: Current access paths for X, Facebook, Instagram, TikTok, Reddit, YouTube, Bluesky, Telegram, and more, including optional Apify scraper integration with shell-safety discipline and evidence archiving requirements. - Use Case: A journalist sees the same claim pushed by dozens of new X accounts. Use this Skill to score each account's authenticity, document coordination signals, archive the posts before deletion, and record findings in the case file with confidence rationale. ## Quick Start Use the social-media-intelligence skill to assess whether these five X accounts pushing the same claim are authentic or part of a coordinated campaign, and archive the key posts.

Frequently Asked Questions about social-media-intelligence

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I detect coordinated inauthentic behavior on social media?▼

Check timing patterns (synchronized posting, burst activity), content patterns (identical text or typos across accounts), account patterns (similar creation dates and usernames), and network patterns (dense clusters amplifying the same sources). Four or more signals indicate strong coordination; treat the accounts as a single source.

How to check if a social media account is a bot or fake?▼

Score the account against red flags: created under 30 days ago, follower/following ratio below 0.1, over 50 posts per day, stock or AI-generated profile photo, and mostly reshared content. Three or more red flags warrant low confidence on any finding sourced from that account.

What tools track how a claim spreads across platforms?▼

Use the Wayback Machine CDX API to find the earliest instance, then map each appearance with platform, author, timestamp, archive URL, and engagement counts. Hoaxy2 visualizes claim spread, and CooRTweet analyzes coordinated posting behavior.

Can I scrape X (Twitter) data for research after the API changes?▼

The official X API is pay-per-use since February 2026, and post-2023 ToS prohibits scraping even public posts. Apify-based collection is possible but ToS-violating; where legal scrutiny is likely, prefer the official API or a licensed broker and record the collection authority.

Why archive social media posts before citing them?▼

Social content disappears through deletion, account suspension, or expiring formats like Instagram Stories. Archive the post URL and profile page via web archiving, download ephemeral video immediately, and record engagement counts at access time since they change.

Does CrowdTangle still work for Facebook research?▼

No, CrowdTangle shut down on August 14, 2024. Meta's replacement is the Meta Content Library and Library API, with applications going through Meta's portal; eligibility favors academic and nonprofit affiliations.