Skip to content
Hugin
Trust - Methodology

Evidence, not accusations.

Hugin turns public material into reusable reports with visible source context, explicit limits, stable URLs, and exportable data.

Public materialNo private account access.

Hugin reads public URLs, public posts, public records, and evidence packets supplied by tools that already hold public material.

Evidence limitsSignals are not accusations.

Reports show patterns, timing, links, and source context. They do not identify a real person or decide misconduct.

AI boundaryModels summarize; code anchors.

AI can name a likely playbook and explain visible evidence. Deterministic checks, source rows, and exports remain inspectable.

What can enter

Public evidence only

  • Public website and Reddit URLs
  • Visible post/profile evidence packets
  • Domain, DNS, TLS, archive, and public threat-intel rows
  • Source receipts for civic/news records
What stays out

Private material stays private

  • No passwords, session cookies, or private account access
  • No scraping behind logins or paywalls
  • No criminal conclusions or identity claims
  • No hidden source substitution when a source is blocked
Machine contract

v1

83 capabilities, 63 endpoints, and 12 limit buckets.

  • GET /api/v1
  • GET /api/v1/status
  • GET /api/v1/limits
  • GET /api/v1/resources
CorrectionsEvery report is disputable at a stable URL.

If Hugin misreads a signal, corrections should point at the exact report, source row, or public record. The evidence trail stays visible so review is grounded in the same packet.