Security audit

SkillMarket

Security checks for vulnerabilities and agentic risk

Overview

Review before installing: this skill openly acts as a marketplace installer, but it can fetch changing third-party install instructions that may alter the agent's installed skills.

Install only if you are comfortable trusting the Zerone marketplace and the individual skill publishers. Before approving any install, read the exact commands, prefer signed or checksummed packages, avoid privileged execution, and use a sandbox or restricted environment where possible.

A.I.G

Vulnerability Patterns

Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

Findings (1)

T03 · Remote Payload Retrieval and Execution

Warning

Location: SKILL.md:45
Finding: Remotely Mutable Installation Instructions May Lead to Untrusted Code Execution
Content: View full analysis
Remediation: View remediation

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (6)

Undeclared Tool Scope

Medium

Category: MCP Least Privilege
Confidence: 91% confidence
Finding: The skill performs network-backed discovery and installation actions but does not declare any explicit tool scope or allowed-tools boundary. That mismatch is dangerous because an agent may invoke network-capable scripts and process untrusted marketplace content without a clear permission model, increasing the chance of overbroad execution and supply-chain abuse.
Content: No source excerpt is available for this finding.

External Transmission

Medium

Category: Data Exfiltration
Confidence: 88% confidence
Finding: The skill is explicitly designed to contact an external marketplace endpoint to discover and install new skills, which creates a real external data transmission and supply-chain trust boundary. Because the returned installation tutorial may influence later command execution or filesystem writes, compromise of the endpoint or malicious marketplace content could lead to installation of untrusted code or exfiltration through follow-on actions.
Content: Scanner excerpt · SKILL.md (reported line 3)May include surrounding context.
md
--- name: skill-market description: Use this skill to find, explore, and install new skills from the Zerone Skill Market (https://api.zerone.market/api). Trigger this when the user asks to "add a skill", "install a skill", "browse skills", or mentions a skill name that is not currently installed. --- # Skill Market

External Transmission

Medium

Category: Data Exfiltration
Confidence: 88% confidence
Finding: This second reference reinforces that the skill is configured around a fixed external API endpoint, so the risk is not incidental but core to operation. In the context of a skill installer, external content is especially dangerous because it can directly influence what gets written into .agent/skills/ and what commands a user may be asked to approve.
Content: Scanner excerpt · SKILL.md (reported line 59)May include surrounding context.
md
Only after explicit user approval, execute the instructions to install the skill into the `.agent/skills/` directory. ## Current Environment Context - **Market Endpoint:** `https://api.zerone.market/api` - **Default Language:** `en` (English is used for API requests as per instructions) - **Local Skills path:** `.agent/skills/`

External Transmission

Medium

Category: Data Exfiltration
Confidence: 60% confidence
Finding: Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
Content: Scanner excerpt · scripts/market.py (reported line 6)May include surrounding context.
python
import urllib.request import urllib.parse ENDPOINT = "https://api.zerone.market/api" def fetch_json(url): try:

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 91% confidence
Finding: The markdown explicitly states 'Default Language: en' and says English is used for API requests as per instructions. This imposes a language choice in the skill behavior without offering the user a locale option or documenting a justified region-specific constraint.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The code hard-codes lang=en in every API call, which imposes a specific language setting on users with no opt-in or alternative. This matches the language/locale policy concern because the skill does not offer locale selection or explain why English is required.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.