Claude 24f816b6a3
Consolidate 22 sibling repos into layered organism structure
Place useful parts of the surrounding repos into sica-fondt by layer, per the
body model (Ada = membrane; brain/endocrine/capabilities/knowledge non-Ada):

- brain/        LLM reasoning + providers (dapr, hermes, MoMoA)
- capabilities/ REPRAG sidecars: hermes tools/skills, dapr tools, parallel
                dispatch, A51 channels, and the OSINT cluster
- knowledge/    LORAG corpus: 754 cyber-skills, agency personas, secure-coding,
                MITRE ATT&CK data
- reference/    defensive threat-reference (C3, shhbruh doc) + AdaYaml parser

License handling: AGPL sources (worldosint, advanced_evolution, mercury,
Reticulum) and GPL DeTTECT are SPEC-only clean-room/port descriptions — no
copyleft code copied. MIT/Apache/data parts copied as working trees.

Safety: shhbruh escape/persistence material and C3 covert-C2 kept as reference
only, not wired into the running organism. See CONSOLIDATION.md.

https://claude.ai/code/session_01UehUqEXXJJCsHoA4voCU5c
2026-06-10 06:53:01 +00:00

1.4 KiB

Workflows: Google Workspace Phishing Protection

Workflow 1: Gmail Inbound Protection Pipeline

Inbound email arrives at Gmail
  |
  v
[Connection-level checks]
  +-- IP reputation
  +-- SPF validation
  +-- DKIM verification
  +-- DMARC evaluation
  |
  v
[Enhanced Pre-Delivery Scanning]
  +-- Content analysis for phishing indicators
  +-- URL expansion (shortened URLs)
  +-- Image scanning for embedded phishing
  +-- NLP analysis for social engineering
  |
  v
[Attachment Protection]
  +-- File type analysis
  +-- Script detection in attachments
  +-- Encrypted attachment from untrusted sender check
  +-- Security Sandbox detonation (Enterprise)
  |
  v
[Spoofing Detection]
  +-- Domain name similarity check
  +-- Employee name impersonation check
  +-- Internal domain spoofing check
  |
  v
[Delivery decision]
  +-- DELIVER: Clean message to inbox
  +-- WARN: Deliver with yellow warning banner
  +-- SPAM: Route to spam folder
  +-- QUARANTINE: Hold for admin review
  +-- REJECT: Block delivery entirely

Workflow 2: Safe Browsing URL Protection

User clicks URL in Gmail
  |
  v
[Enhanced Safe Browsing check]
  +-- Real-time URL reputation lookup
  +-- Check against known phishing database
  +-- Dynamic page analysis
  |
  v
[Decision]
  +-- SAFE: Allow navigation
  +-- DANGEROUS: Display full-page warning
  +-- SUSPICIOUS: Display interstitial warning