Trinity Search

Search millions of public records with four-state structural results. Every query tells you what's affirmed, denied, mentioned, and absent.

OSHA
FDA FAERS
CVE/NVD
FEC
Exploit-DB
--
Public Records
5
Databases
4
Search States
$0
Cost to Search

Search Examples

Click any example to search instantly. Comma = AND query.

OSHA Severe Injury Reports

102,921 workplace injuries. Search by injury type, body part, state, industry, or employer.
Affirmed = Hospitalized / Amputation
Absent = No such injury in records

By Injury Type

amputation fracture burn electrocution crush laceration fall

By Body Part

hand finger arm leg head eye foot back

By State

Texas California Ohio Florida Pennsylvania Illinois New York

By Industry

construction manufacturing warehouse restaurant roofing

Compound Searches (AND queries)

amputation + Texas burn + chemical fall + construction crush + manufacturing hand + amputation struck + warehouse

By Employer

Amazon Walmart Tyson Tesla

FDA Adverse Event Reports (FAERS)

351,700 drug safety reports. Search any drug to see its death/serious/mild breakdown instantly.
Affirmed = Death or Life-Threatening
Mentioned = Hospitalized / Disability
Denied = Other / Mild outcome

Drug Safety Profile

Instant severity breakdown for any drug.
Ozempic Xarelto Metformin Humira Aspirin Eliquis Keytruda Wegovy

Drug + Reaction

Does this drug cause this side effect?
Ozempic + nausea Ozempic + pancreatitis Xarelto + bleeding Metformin + lactic acidosis Aspirin + ulcer Humira + infection

Reaction Landscape

Which drugs cause this reaction?
cardiac hepatic (liver) renal (kidney) anaphylaxis nausea pneumonia seizure

Demographic

Ozempic in women Ozempic in men Cardiac in women

CVE/NVD Vulnerability Database

331,723 software vulnerabilities. Instant severity breakdown by vendor, type, or product.
Affirmed = CRITICAL / HIGH
Mentioned = MEDIUM
Denied = LOW

By Vendor

Severity breakdown for any vendor.
Microsoft Apple Google Linux Apache Oracle Cisco Adobe

By Vulnerability Type

injection overflow XSS authentication remote privilege denial of service

Vendor + Severity

Microsoft CRITICAL Apache CRITICAL Linux HIGH Google CRITICAL

Type + Attack Vector

Remote injection Remote overflow Remote XSS Local privilege escalation

By Product

Chrome Windows Android WordPress Nginx

FEC Campaign Contributions

58M+ individual donations (2024 cycle). Search by employer, occupation, or state.
Mentioned = Contribution recorded
Absent = No matching contributions

By Employer

Which companies' employees are donating?
Google Amazon Walmart Microsoft Goldman Sachs Apple

By Occupation

Attorneys Engineers Retired Physicians CEOs Professors

Employer + State

Google in California Amazon in Washington Goldman in New York

Exploit-DB Code Analysis

46,487 exploit source files safely decomposed into searchable elements. Code is inert — executable form permanently destroyed.
Affirmed = CRITICAL / HIGH threat
Mentioned = MEDIUM threat
Denied = LOW / INFO

By Danger Pattern

Find exploits using specific attack techniques
Shellcode Reverse Shell Buffer Overflow Injection Privilege Escalation Ransomware Keylogger DLL Injection

By Language

Python C/C++ PHP Ruby Shell/Bash JavaScript Assembly

By Platform

Windows Linux Android iOS Multi-platform

By Exploit Type

Remote Local Web Application Denial of Service

By Import / Library

What libraries do exploits use?
Socket Subprocess Struct ctypes HTTP requests Base64

Compound Searches

Windows shellcode Linux buffer overflow PHP injection Python reverse shells Linux privesc Ransomware + encryption

Cross-Database Intelligence

Correlations across datasets that were never searchable together before.

Corporate Safety vs Donations

Companies with workplace injuries — are they also major political donors?
OSHAFEC
Amazon Walmart Tyson Tesla

Drug Safety vs Lobbying

Drugs with fatal adverse events — how much are their manufacturers donating?
FAERSFEC
Ozempic Xarelto Humira Keytruda

Tech Donors & Cyber Exposure

Tech companies donating politically vs their published software vulnerabilities.
FECCVE
Cisco Dell Samsung Adobe

Cyber + Physical Safety

Companies with software vulnerabilities AND workplace injuries.
CVEOSHA
Amazon Siemens Honeywell Apple

Corporate Trifecta

Companies across safety, cyber, and political spending — three independent dimensions at once.
OSHACVEFEC
Amazon Siemens Honeywell

Vulnerability → Exploit Pipeline

When a CVE is disclosed vs when exploit code targeting it appears in the wild.
CVEExploits
Apache WordPress Microsoft Privilege Escalation

Workplace Exposure & Drug Reactions

Chemical injuries at work vs related FDA adverse events. Same substances, different contexts.
OSHAFAERS
Chemical Burn Respiratory Fracture

Computational Intelligence

Not just search — computation over the compressed data. Zero new storage. Pure structural algebra.

Co-Occurrence Analysis

What words appear alongside a term across thousands of documents? Reveals hidden patterns via PMI.
OSHAFAERSFEC
FENTANYL (FAERS) apache (Exploits) amputation (OSHA) DEATH (FAERS)

Vocabulary Profile

Aggregate word statistics across all documents matching a query. Top terms, frequency distribution, document density.
OSHACVEExploits
OZEMPIC (FAERS) apache (Exploits) construction (OSHA) remote (CVE)

Term Transform

Find-and-replace at the mathematical level. Swap any term, see the transformed output — no data is changed.
Any DB
amputation → fracture (OSHA) nausea → vomiting (FAERS)

Document Similarity

Jaccard, cosine, or Dice similarity between any two documents. Pure mathematical comparison on compressed representations.
Any DB
Find similar (OSHA) Find similar (FAERS)

Anomaly Detection

Find documents that deviate most from the statistical center of a query result set. Automated outlier scoring.
OSHAFAERSCVE
amputation outliers (OSHA) OZEMPIC outliers (FAERS) remote outliers (CVE)

Cross-Database Similarity

Find similar documents across independent databases. Cross-database matching performed automatically at query time.
OSHA→FAERSCVE→Exploits
OSHA → FAERS CVE → Exploits

Phrase Search

Find documents containing exact phrases. Instant phrase matching across millions of records.
OSHAFAERSExploits
struck by (OSHA) cardiac arrest (FAERS) fell from (OSHA)

Temporal Co-Occurrence

Date-filtered co-occurrence analysis. See how word relationships shift across time periods.
OSHACVEFAERS
amputation 2022-23 (OSHA) amputation 2015-17 (OSHA) remote 2023-24 (CVE)

Four Answers to Every Question

No other search system on earth produces this.

Affirmed

The document contains your search terms AND the data confirms it. The adverse event happened. The vote was cast. The vulnerability is critical.

Denied

The document contains your terms BUT the data says NO. The condition was screened and ruled out. The vote was against. The severity is low.

Mentioned

Your terms appear in the document but no polarity is declared. It's there, but the data doesn't take a position on it.

Absent

Your terms do not appear in this corpus. This isn't "no results found" — it's a definitive finding. The absence IS the answer.

The Databases

All public records. All searchable. All in one place.

Life & Safety

OSHA Severe Injury Reports

U.S. Occupational Safety & Health Administration
Every reported workplace amputation, hospitalization, and fatality.
--Records
2015–2025Coverage
Life & Safety

FDA Adverse Event Reports

FDA FAERS
Drug side effects, hospitalizations, and deaths reported to the FDA.
--Reports
Q3 2024Coverage
Life & Safety

CVE/NVD Vulnerabilities

NIST National Vulnerability Database
Every known software vulnerability with CVSS severity scores.
--CVEs
1999–2025Coverage
Rights & Accountability

FEC Campaign Contributions

Federal Election Commission — 2024 Cycle
Every individual campaign contribution over $200.
--Contributions
2024Cycle
Life & Safety

Exploit-DB Code Analysis

Exploit Database — Safely Decomposed
Exploit source code decomposed into inert, searchable form. Executable structure permanently destroyed.
--Exploits
2004–2025Coverage

Three Principles

1. The Absence Principle

Trinity creates a complete index. If a term isn't found, no document contains it. "Not found" means "definitively does not exist."

2. The Exclusion Principle

Documents don't just contain terms — they assert or deny them. Structural polarity markers distinguish affirmation from denial at the data level.

3. Custom Search Models

Different users see different things in the same data. Models are query-time overlays — zero data duplication.