AgentPit 财经综合分析

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:41 PM UTC
• Version:
  1.0.0

Overview

This skill is a disclosed paid AgentPit stock-analysis integration that needs an API-style key and makes user-selected remote analysis calls.

Install only if you trust AgentPit and are comfortable storing its consumerKey locally for reuse. Confirm which analysis systems should run before approving calls, monitor AgentPit billing, and delete ~/.openclaw/secrets/agentpit.cpk plus unset AGENTPIT_CPK if you want to revoke local access.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (2)

Description-Behavior Mismatch

Medium

• Confidence: 95% confidence
• Finding: The skill instructs the agent to collect a user consumer key and persist it to a local secrets file, which expands the skill from stock analysis into credential handling and long-term secret storage. Persisting third-party API credentials increases the blast radius of any host compromise, accidental disclosure, or later misuse by unrelated skills or processes.

Context-Inappropriate Capability

Medium

• Confidence: 92% confidence
• Finding: The documentation directs the agent to manage local credential files, environment variables, and permissions even though the declared purpose is financial analysis. This unnecessary privilege expansion creates avoidable opportunities for secret persistence, leakage to subprocesses, and abuse by other local components that can read environment or filesystem state.

VirusTotal

65/65 vendors flagged this skill as clean.