AgentReadyHomeAgent Listing

← Glm Image

Glm Image — agentic threat model

6.0AIVSS 6.0 · Medium

GLM Image is primarily a text-to-image generation model accessed via APIs rather than an autonomous agent, presenting low agentic risk. Its primary security concerns center on model-level vulnerabilities (adversarial prompts, output alignment) and API security rather than autonomous execution threats.

OWASP AIVSS score rationale

AIVSS = (CVSS_Base + AARS) × Mitigation_Factor, where AARS = (10 − CVSS_Base) × (Factor_Sum / 10) × ThM
CVSS base 5.3AARS uplift 0.67Factor sum 1.5/10Threat ×0.95Mitigation ×1.0
Autonomy of Action
0.10
Goal-Driven Planning
0.00
Self-Modification
0.00
Dynamic Tool Use
0.00
Persistent Memory
0.00
Contextual Awareness
0.20
Dynamic Identity
0.00
Multi-Agent Interactions
0.00
Non-Determinism
0.50
Opacity & Reflexivity
0.70

Scored with the canonical OWASP AIVSS formula (AIVSS calculator reference); agentic risk factors estimated from the agent’s described capabilities.

MAESTRO 7-layer threat model

Per-layer threats for this agent. Layers tagged “not certain from listing” are general, caveated commentary where the public description didn’t pin that layer.

L1 · Foundation Models✓ mapped

Utilizes a hybrid 9B autoregressive core and a 7B DiT diffusion decoder. Primary threats include adversarial prompt injection to bypass safety filters, model extraction/stealing via API querying, and generation of misaligned or harmful outputs.

L2 · Data Operations⚠ not certain from listing

Not certain from the listing — training data details are proprietary. Threats include potential copyright infringement in training data, data poisoning during pre-training, and lack of transparency in data lineage.

L3 · Agent Frameworks⚠ not certain from listing

Not certain from the listing — GLM Image is primarily a model/API rather than an active agent framework. If integrated into orchestrators, threats include insecure tool integration and prompt injection via orchestrating frameworks.

L4 · Deployment & Infrastructure⚠ not certain from listing

Not certain from the listing — hosted as a cloud API with Python/Java SDKs. Threats include API key exposure, unauthorized API access, and potential denial of service on the generation endpoints.

L5 · Evaluation & Observability⚠ not certain from listing

Not certain from the listing — no built-in guardrails or observability tools are mentioned. Threats include a lack of input/output filtering for NSFW or policy-violating generations.

L6 · Security & Compliance (cross-cutting)⚠ not certain from listing

Not certain from the listing — compliance certifications (like SOC2 or GDPR) are not specified. Threats include compliance violations if used to generate copyrighted or deepfake material without consent.

L7 · Agent Ecosystem⚠ not certain from listing

Not certain from the listing — does not natively interact in a multi-agent ecosystem. Threats are limited to downstream applications consuming its API without proper validation.

MAESTRO — the 7-layer agentic threat-modeling framework (Cloud Security Alliance / Ken Huang).