CtrlK
BlogDocsLog inGet started
Tessl Logo

multi-agentic-threat-model

Comprehensive threat modeling for multi-agent systems using CSA MAESTRO 7-layer framework and OWASP Multi-Agentic System Threat Modeling Guide v1.0. Systematically analyzes threats across all architectural layers from foundation models to agent ecosystems.

56

1.61x
Quality

33%

Does it follow best practices?

Impact

97%

1.61x

Average score across 3 eval scenarios

SecuritybySnyk

Passed

No known issues

Optimize this skill with Tessl

npx tessl skill review --optimize ./skills/multi-agentic-threat-model/SKILL.md
SKILL.md
Quality
Evals
Security

Evaluation results

100%

54%

Healthcare AI Triage System Security Analysis

Full MAESTRO 7-layer threat model output structure

Criteria
Without context
With context

Layer 1 named

0%

100%

Layer 2 named

0%

100%

Layer 3 named

0%

100%

Layer 4 named

0%

100%

Layer 5 named

0%

100%

Layer 6 named

0%

100%

Layer 7 named

0%

100%

Architecture map section

62%

100%

Layer-specific threat section

37%

100%

Cross-layer attack section

100%

100%

Extended threat section

75%

100%

Architecture pattern section

100%

100%

Risk matrix section

100%

100%

Mitigation section

100%

100%

CSA/MAESTRO attribution

0%

100%

93%

48%

Autonomous Financial Trading Platform Threat Analysis

Extended multi-agent threats and architecture pattern risks

Criteria
Without context
With context

Reasoning Collapse identified

0%

100%

Emergent Covert Coordination identified

25%

100%

Heterogeneous Multi-Agent Exploits identified

0%

100%

Goal Drift identified

12%

100%

Trust Misuse identified

12%

100%

Supervisor-agent pattern assessed

100%

100%

Hierarchical pattern assessed

62%

62%

Distributed ecosystem pattern assessed

25%

50%

Human-in-the-loop pattern assessed

100%

100%

Risk matrix present

100%

100%

threat_summary.json produced

100%

100%

MAESTRO framework referenced

0%

100%

100%

11%

DevOps Automation Platform Attack Path Analysis

Cross-layer attack chains and layered mitigation strategy

Criteria
Without context
With context

Supply chain attacks covered

100%

100%

Lateral movement covered

100%

100%

Privilege escalation covered

100%

100%

Data leakage cascade covered

100%

100%

Multi-layer propagation described

100%

100%

Layer-specific mitigations

90%

100%

Cross-layer mitigations

100%

100%

AI-specific mitigations

100%

100%

MAESTRO layer names used

0%

100%

attack_chains.md produced

100%

100%

mitigations.md produced

100%

100%

Repository
OWASP/secure-agent-playbook
Evaluated
Agent
Claude Code
Model
Claude Sonnet 4.6

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.