OSINT & Due Diligence

Adverse Media Sweep

Scan every public surface — press, web, social, video, and forums — for lawsuits, fraud claims, and negative chatter before you onboard a person or company

What it does

KYC/EDD teams need negative-coverage evidence, not vibes. This chains dated press wires with legal-keyword web queries and sentiment-filtered Reddit, X, YouTube, and forum chatter — and lets you toggle which surfaces run and pin a jurisdiction — so red flags surface even when they never made the news.

Who it's for: Compliance, KYC/AML, and investor due-diligence teams

Inputs

How your agent runs it

1. 1
   search_news

   search_news(query="{subject_name} {context_terms}", time_published=1y, limit=50)

   Backbone (always run): pull the last year of press coverage and keep articles framed around legal, regulatory, or financial-misconduct language. Pass country={country} only when screening a specific jurisdiction.

2. 2
   search_web

   search_web(query="{subject_name} lawsuit OR fraud OR investigation OR settlement OR scam", time=year)

   Backbone (always run): surface court filings, regulator notices, and complaint sites that never reach mainstream wires. Pass country={country} only when screening a specific jurisdiction.

3. 3
   search_reddit

   search_reddit(query="{subject_name}", sort_by=top, get_sentiment=true)

   Only run if {platforms} includes reddit: keep top threads whose polarity is negative or dominant_emotion is anger/disgust to catch customer or investor grievances.

4. 4
   search_twitter

   search_twitter(query="{subject_name}", sort_by=most_recent, pages=5, get_sentiment=true)

   Only run if {platforms} includes twitter: flag recent negatively-polarized mentions and complaint threads, noting accounts repeating the same allegation.

5. 5
   search_youtube

   search_youtube(query="{subject_name} scam OR fraud OR lawsuit OR exposed", upload_date=this_year, get_sentiment=true)

   Only run if {platforms} includes youtube: capture investigative or expose videos and scam warnings; retain negatively-toned clips and note channels that show corroborating documents or filings.

6. 6
   search_forums

   search_forums(query="{subject_name}", time=year, get_sentiment=true)

   Only run if {platforms} includes forums: capture niche-forum scam reports and warnings, retaining only negative-polarity posts with corroborating detail. Pass country={country} only when screening a specific jurisdiction.

Tools used

The full playbook

This is exactly what the MCP returns to your agent (via the adverse-media-sweep prompt or get_skill tool), with your inputs filled in.

SKILL: Adverse Media Sweep
KYC/EDD teams need negative-coverage evidence, not vibes. This chains dated press wires with legal-keyword web queries and sentiment-filtered Reddit, X, YouTube, and forum chatter — and lets you toggle which surfaces run and pin a jurisdiction — so red flags surface even when they never made the news.

You are running this skill on API Direct via its MCP tools. Execute the steps below yourself by calling the named tools in order — values in <angle brackets> come from a previous step. Then deliver the result described at the end.

INPUTS:
  - subject_name (required): <subject_name — ASK THE USER>
      The person or legal entity being screened
  - context_terms (optional): (optional — e.g. crypto fund Miami)
      Disambiguating context (industry, city, or aliases) to avoid same-name false positives
  - platforms (optional): (optional — e.g. reddit, twitter, youtube, forums)
      Comma-separated list of platforms to run — options: news, web, reddit, twitter, youtube, forums. Omit to run all of them; name specific platforms to limit the run.
  - country (optional): (optional — e.g. us)
      ISO country code to focus jurisdiction-specific news, web, and forum results (maps to the real country param on those tools). Omit for a global sweep.

PLATFORM SELECTION: some steps are gated with "Only run if {platforms} includes X". If the user supplied a `platforms` value, run only the steps whose platform is listed. If the user did NOT supply `platforms`, run ALL steps — every platform the skill supports.

STEPS:
  1. Tool `search_news` — search_news(query="{subject_name} {context_terms}", time_published=1y, limit=50)
     Backbone (always run): pull the last year of press coverage and keep articles framed around legal, regulatory, or financial-misconduct language. Pass country={country} only when screening a specific jurisdiction.
  2. Tool `search_web` — search_web(query="{subject_name} lawsuit OR fraud OR investigation OR settlement OR scam", time=year)
     Backbone (always run): surface court filings, regulator notices, and complaint sites that never reach mainstream wires. Pass country={country} only when screening a specific jurisdiction.
  3. Tool `search_reddit` — search_reddit(query="{subject_name}", sort_by=top, get_sentiment=true)
     Only run if {platforms} includes reddit: keep top threads whose polarity is negative or dominant_emotion is anger/disgust to catch customer or investor grievances.
  4. Tool `search_twitter` — search_twitter(query="{subject_name}", sort_by=most_recent, pages=5, get_sentiment=true)
     Only run if {platforms} includes twitter: flag recent negatively-polarized mentions and complaint threads, noting accounts repeating the same allegation.
  5. Tool `search_youtube` — search_youtube(query="{subject_name} scam OR fraud OR lawsuit OR exposed", upload_date=this_year, get_sentiment=true)
     Only run if {platforms} includes youtube: capture investigative or expose videos and scam warnings; retain negatively-toned clips and note channels that show corroborating documents or filings.
  6. Tool `search_forums` — search_forums(query="{subject_name}", time=year, get_sentiment=true)
     Only run if {platforms} includes forums: capture niche-forum scam reports and warnings, retaining only negative-polarity posts with corroborating detail. Pass country={country} only when screening a specific jurisdiction.

DELIVER: A consolidated adverse-media memo ranking negative findings by severity, each tied to a dated source link (press, web, social, or video) for the KYC/EDD file

Note: each underlying tool call is billed at its normal endpoint price; get_sentiment adds a small per-page surcharge. Page through results as needed but stop once you have enough to deliver the outcome.