Security audit

Sharebench

Security checks for vulnerabilities and agentic risk

Overview

This skill helps agents search a public registry for reusable AI artifacts and does not install code, persist data, or request privileged access.

Installers should treat Sharebench results as external reference material: review any retrieved artifact before using it, do not let remote SKILL.md text override local instructions, avoid sensitive search terms, and get explicit confirmation before installing or running anything found through the registry.

A.I.G

Vulnerability Patterns

Insecure DependenciesIntroduces malicious components through unsafe dependency sources
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
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally

Findings (1)

T08 · Insecure Dependencies

Warning

Location: SKILL.md:35
Finding: Untrusted Remote AI Artifact Supply Chain
Content: View full analysis
``` ``` The optional MCP integration further introduces an external service endpoint: ```markdown ## MCP option (for MCP-capable clients) If your client speaks the Model Context Protocol, you can connect the Sharebench server directly for the same search-and-read capability as tools: ``` https://mcp-public.sharebench.ai ``` ``` ### Technical Analysis The Skill instructs an agent to retrieve community-contributed AI artifacts from an external registry and consider applying their contents. Those artifacts are mutable data outside the audited package and therefore fall outside the trust boundary established by this review. Although the document advises the agent to read artifacts fully, preserve attribution, and prefer visible provenance, it does not impose enforceable security controls requiring the agent to: - Treat retrieved `SKILL.md` content strictly as untrusted reference data. - Ignore instructions that attempt to override system, developer, user, or local Skill directives. - Verify an immutable artifact version or cryptographic dige ...[truncated 2878 chars]
Remediation: View remediation

SkillSpector

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (2)

Ae1

High

Category: analysis-evasion
Confidence: 100% confidence
Finding: Referenced artifact was not completely inspected
Content: Scanner excerpt · SKILL.md (reported line 12)May include surrounding context.
md
`SKILL.md` format, contributed by teams and the community. Every public

Ae1

High

Category: analysis-evasion
Confidence: 100% confidence
Finding: Referenced artifact was not completely inspected
Content: Scanner excerpt · SKILL.md (reported line 54)May include surrounding context.
md
`SKILL.md` format, contributed by teams and the community. Every public