osint-investigation

Cross-reference public-record datasets and generate confidence-scored evidence chains.

1|Updated May 16, 2026
One-click install
npx skills add https://github.com/devMoez/titan --skill osint-investigation-devmoez
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: osint-investigation
Source: https://github.com/devMoez/titan/tree/main/optional-skills/research/osint-investigation
Command: npx skills add https://github.com/devMoez/titan --skill osint-investigation-devmoez

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

OSINT investigations become unreliable when evidence is scattered across inconsistent sources and entity identities don’t match cleanly. This Skill helps you systematically cross-reference public records, track confidence, and assemble a verifiable evidence chain from raw source rows.

Core Features & Use Cases

  • Cross-source OSINT investigation: Pulls from major public-record datasets like SEC EDGAR, USAspending, Senate lobbying disclosures, OFAC sanctions, ICIJ offshore leaks, NYC ACRIS property records, CourtListener litigation data, Wayback archives, Wikipedia/Wikidata, and GDELT news monitoring.
  • Entity resolution with confidence tiers: Normalizes names and produces candidate links (exact, fuzzy, and token-overlap) with explicit confidence—so matches are leads, not unsupported conclusions.
  • Evidence-chain JSON output: Produces structured findings where every claim ties back to specific CSV rows from the underlying sources.
  • Optional timing correlation tests: Runs a permutation test to see whether events (e.g., lobbying filings) cluster suspiciously near related counterpart events (e.g., contract awards), without claiming wrongdoing.

Quick Start

Use the skill to investigate a company’s public-record footprint by having the agent run the cross-reference workflow and return a findings.json with confidence-scored evidence chains and optional timing analysis.

Frequently Asked Questions about osint-investigation

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

FAQPage Schema
How do I build a defensible public-record evidence chain for due diligence?

To build a defensible public-record evidence chain, you cross-reference heterogeneous datasets like filings, sanctions, and court cases, generating JSON output that links every finding to specific source rows for traceability.

What is the best way to resolve entity identities across inconsistent OSINT datasets?

Entity resolution across inconsistent OSINT datasets is handled by normalizing names and producing candidate links using exact, fuzzy, and token-overlap confidence tiers, ensuring matches serve as leads rather than unsupported conclusions.

Can I perform sanctions screening and cross-reference public records using only Python stdlib?

Yes, you can perform sanctions screening and cross-reference public records using only Python stdlib fetchers, which collect data from OFAC, SEC EDGAR, and property records to output normalized CSV files.

Does OSINT investigation require traceability rather than narrative guesswork for compliance?

OSINT investigation for compliance requires traceability rather than narrative guesswork, applying cross-referencing across filings and court cases to produce structured findings where every claim ties back to specific CSV source rows.

How do I test if lobbying filings and contract awards cluster suspiciously without claiming wrongdoing?

To test if lobbying filings and contract awards cluster suspiciously, you run optional timing correlation permutation tests that check whether events cluster near counterpart events without claiming wrongdoing.

What public-record sources are supported for cross-source entity resolution and sanctions screening?

Supported public-record sources for cross-source entity resolution and sanctions screening include SEC EDGAR, USAspending, Senate lobbying disclosures, OFAC, ICIJ offshore leaks, NYC ACRIS, CourtListener, Wayback archives, and GDELT.