ClawHub
Back to skill

Security audit

Chat Value Insight 会话价值洞察

Security checks across malware telemetry and agentic risk

Overview

This skill locally analyzes user-provided chat exports and its sensitive behavior is mostly disclosed, scoped, and purpose-aligned.

Install only if you are comfortable giving the skill access to the chat export paths you explicitly provide and with generated reports containing excerpts from those chats. Be aware that setup/report output may include whether related skills or optional ASR API-key environment variables are present, though the artifact does not read key values or call cloud ASR in the core path.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Lp3

Medium
Category
MCP Least Privilege
Confidence
82% confidence
Finding
The skill advertises and instructs use of file reads/writes, shell execution, and environment-variable-aware behavior, but the metadata does not declare corresponding permissions. This creates a transparency and policy-enforcement gap: an agent or platform may execute capabilities users did not explicitly approve, especially when processing sensitive chat exports containing private conversations and local file paths.

Context-Inappropriate Capability

Medium
Confidence
95% confidence
Finding
The function enumerates unrelated sibling skill directories and reports whether specific skills are installed, plus whether certain API key environment variables are present. For a chat-export analysis skill, this host reconnaissance is not necessary for core functionality and creates privacy-sensitive telemetry about the user's local toolchain and secret availability, which could later be exposed in logs, prompts, or reports.

Env Variable Harvesting

High
Category
Data Exfiltration
Content
for n in names:
        out[n] = any((r / n).exists() for r in roots)
    out["dashscope_key_present"] = bool(
        os.environ.get("DASHSCOPE_API_KEY") or os.environ.get("BAILIAN_API_KEY")
    )
    return out
Confidence
88% confidence
Finding
os.environ.get("DASHSCOPE_API_KEY

Env Variable Harvesting

High
Category
Data Exfiltration
Content
for n in names:
        out[n] = any((r / n).exists() for r in roots)
    out["dashscope_key_present"] = bool(
        os.environ.get("DASHSCOPE_API_KEY") or os.environ.get("BAILIAN_API_KEY")
    )
    return out
Confidence
88% confidence
Finding
os.environ.get("BAILIAN_API_KEY

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.