/02. aio-threat-models

PLUGINS

Install: /plugin install aio-threat-models@aiocean-plugins · v1.0.2

aio-threat-models

Threat modeling knowledge advisor with semantic search across 27 frameworks — STRIDE, LINDDUN, PASTA, OCTAVE, Attack Trees, MITRE ATT&CK/ATLAS, Kill Chain, NIST AI RMF, OWASP LLM Top 10, and more.

What This Is

A knowledge skill (à la aio-mental-models) that:

  1. Catalogs 27 threat modeling frameworks across 7 volumes — each with its own markdown file covering origin, structure, when-to-use, strengths, limitations, cross-framework relations, and primary-source citations.
  2. Embeds semantic search via pre-computed embeddings (snowflake-arctic-embed-xs, runs locally, no API key).
  3. Guides framework selection, application, and cross-framework composition through a 5-step workflow (ASK → SEARCH → APPLY → COMPOSE → CHALLENGE).

Coverage (7 Volumes, 27 Frameworks)

VolumeFrameworks
1. FoundationsShostack 4Q, DFD + Trust Boundaries, Threat Modeling Manifesto
2. Core FrameworksSTRIDE, DREAD, PASTA, OCTAVE Allegro, Trike, VAST
3. Attacker-centricAttack Trees, Persona non Grata, Security Cards, hTMM
4. Attack PatternsLockheed Martin Kill Chain, Unified Kill Chain, MITRE ATT&CK, Diamond Model, CAPEC
5. PrivacyLINDDUN, GDPR DPIA
6. AI/MLOWASP Top 10 for LLM, MITRE ATLAS, NIST AI RMF, NIST AI 100-2
7. Cloud & DevSecOpsCloud Threat Modeling, Kubernetes Threat Matrix, Continuous Threat Modeling

Triggers

"threat model", "STRIDE", "LINDDUN", "PASTA", "MITRE ATT&CK", "attack surface", "privacy review", "DPIA", "adversarial ML", "secure architecture review", "which framework", "threat modeling for AI", etc.

Scripts

  • list-models.sh — list all 27 frameworks, filter by volume, search by keyword
  • search-models.ts — semantic search (cosine similarity against pre-computed embeddings)
  • build-embeddings.ts — rebuild embeddings.json after content changes

Acknowledgment

Structure and semantic search pattern adapted from aio-mental-models. Content synthesized from primary sources: NIST, CMU/SEI, OWASP, MITRE, Microsoft SDL, Threat Modeling Manifesto, peer-reviewed papers (Scandariato 2015, Tuma 2018, Xiong 2019).

Skills (1)

  • aio-threat-models — Threat modeling framework advisor — selects, applies, and composes frameworks (STRIDE, LINDDUN, PASTA, MITRE ATT&CK/ATLAS, Attack Trees, etc.) for security, pri…

Articles