AgentReadyHomeAgent Listing

← loc-guardian

loc-guardian — agentic threat model

7.4AIVSS 7.4 · High

loc-guardian operates directly on the file-write path as a Claude Code plugin, presenting risks of local file modification, write-gating denial of service, or command injection if the underlying tools or LLM-generated extraction plans are compromised.

OWASP AIVSS score rationale

AIVSS = (CVSS_Base + AARS) × Mitigation_Factor, where AARS = (10 − CVSS_Base) × (Factor_Sum / 10) × ThM
CVSS base 6.5AARS uplift 0.91Factor sum 2.6/10Threat ×1.0Mitigation ×1.0
Autonomy of Action
0.60
Goal-Driven Planning
0.40
Self-Modification
0.00
Dynamic Tool Use
0.30
Persistent Memory
0.10
Contextual Awareness
0.40
Dynamic Identity
0.00
Multi-Agent Interactions
0.00
Non-Determinism
0.50
Opacity & Reflexivity
0.30

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

Uses Claude Opus to generate extraction plans. Threats include prompt injection or adversarial inputs that manipulate the model into generating malicious, corrupted, or syntactically invalid code-splitting plans.

L2 · Data Operations✓ mapped

Reads local codebase files to count lines of code and analyze structure. While it does not use a vector database, malicious local files could exploit parser vulnerabilities in the LOC counter or the LLM context window.

L3 · Agent Frameworks✓ mapped

Integrates with the external tool 'tokei' and runs as a write/edit hook. If file paths or shell executions are not properly sanitized, this integration could lead to local command injection or unauthorized file system traversal.

L4 · Deployment & Infrastructure⚠ not certain from listing

Not certain from the listing — The plugin runs locally within the user's Claude Code environment. Sandboxing and privilege isolation depend entirely on the host system and the parent Claude Code execution environment.

L5 · Evaluation & Observability⚠ not certain from listing

Not certain from the listing — Beyond flagging LOC violations on write, there is no mention of security logging, audit trails, or anomaly detection to monitor if the plugin is bypassed or abused.

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

Not certain from the listing — Enforces a basic compliance policy (LOC limits), but lacks built-in identity, authorization, or access control mechanisms, relying instead on the host operating system's file permissions.

L7 · Agent Ecosystem✓ mapped

Operates as a single-agent plugin within Claude Code. There are no multi-agent interactions or marketplace dependencies described, limiting ecosystem-level cascading risks.

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