AgentReadyHomeAgent Listing

← mcp-trove

mcp-trove — agentic threat model

7.1AIVSS 7.1 · High

mcp-trove acts as a high-value target storing encrypted secrets and snippets; its primary risk is prompt injection on consuming agents leading to unauthorized credential extraction.

OWASP AIVSS score rationale

AIVSS = (CVSS_Base + AARS) × Mitigation_Factor, where AARS = (10 − CVSS_Base) × (Factor_Sum / 10) × ThM
CVSS base 8.5AARS uplift 0.42Factor sum 2.8/10Threat ×1.0Mitigation ×0.8
Autonomy of Action
0.20
Goal-Driven Planning
0.10
Self-Modification
0.00
Dynamic Tool Use
0.40
Persistent Memory
0.80
Contextual Awareness
0.20
Dynamic Identity
0.50
Multi-Agent Interactions
0.30
Non-Determinism
0.10
Opacity & Reflexivity
0.20

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⚠ not certain from listing

Not certain from the listing — mcp-trove is an MCP tool/server rather than a foundation model itself, but it is highly vulnerable to adversarial prompt injections targeting the LLMs that call it.

L2 · Data Operations✓ mapped

Manages a git-backed vault for plaintext snippets and encrypted secrets. Threats include unauthorized data exfiltration of secrets and Git repository poisoning.

L3 · Agent Frameworks✓ mapped

Exposes MCP tools for secret retrieval. Threats include insecure tool integration and indirect prompt injection forcing the host agent to call retrieval tools maliciously.

L4 · Deployment & Infrastructure✓ mapped

Relies on local or remote Git storage and file-system key paths. Threats include local privilege escalation to read the encryption key path or compromise the Git host.

L5 · Evaluation & Observability⚠ not certain from listing

Not certain from the listing — there is no mention of built-in logging, access auditing, or anomaly detection for secret retrieval attempts.

L6 · Security & Compliance (cross-cutting)✓ mapped

Uses key-path based encryption and access controls. Threats include weak key management, lack of robust authentication between the calling agent and the MCP server, and compliance gaps in credential handling.

L7 · Agent Ecosystem✓ mapped

Operates within the MCP ecosystem where other agents can request secrets. Threats include rogue or compromised agents abusing trust to dump credentials horizontally.

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