ClawHub
Back to skill

Security audit

agent-tracer

Security checks across malware telemetry and agentic risk

Overview

The skill is a coherent tracing helper, but it encourages broad raw logging of agent activity to a persistent HTTP service without enough privacy, access-control, or redaction guidance.

Install only if you are prepared to treat the tracer as a sensitive logging system. Keep it bound to localhost or a protected private network, add authentication/TLS if exposed beyond the host, redact secrets and personal data from arguments/results, and define retention/deletion rules before recording real agent runs.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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)

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The skill instructs users to record every tool call, including arguments, results, session identifiers, model names, and token usage, to an HTTP service without any explicit warning about sensitive data handling, minimization, or redaction. In practice, tool arguments and results often contain prompts, retrieved documents, API outputs, secrets, PII, or business data, so this creates a real risk of unintended data exposure even if the service is self-hosted.

External Transmission

Medium
Category
Data Exfiltration
Content
| python3 -c "import sys,json;print(json.load(sys.stdin)['trace_id'])")

# 2. 每次工具调用记一条 span（model/tokens 可选，带上才有成本分析）
curl -s -X POST $TRACER_BASE_URL/traces/$TRACE_ID/spans \
  -H 'Content-Type: application/json' \
  -d '{"tool_name":"web_search","arguments":{"query":"上证指数"},
       "result":"上证 3200.15 +0.85%","duration_ms":850,
Confidence
95% confidence
Finding
curl -s -X POST $TRACER_BASE_URL/traces/$TRACE_ID/spans \ -H 'Content-Type: application/json' \ -d

VirusTotal

54/54 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.