Thrd Skill

• Outcome: Pass
• Latest audit: May 28, 2026 at 12:26 PM UTC
• Version:
  1.2.7

Overview

The skill is a coherent email-inbox integration with disclosed credentials, polling, and optional callback behavior; its main risks are ordinary operational controls, not hidden malicious behavior.

Install this only if you want an agent to operate a dedicated thrd.email inbox. Keep THRD_API_KEY in a secret manager, define clear approval rules for outbound email and billing upgrades, use the default Thrd API endpoint unless you trust the alternative, and only configure --on-events with harmless commands you would run yourself.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Behavioral AST: exec() Call, eval() Call, Dynamic Import
• 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)

subprocess module call

Medium

• Category: Dangerous Code Execution

Content

print(json.dumps(summary))

                if on_events_cmd:
                    subprocess.run(on_events_cmd, shell=False, check=False)

                ack = requests.post(
                    f"{base_url}/v1/events/ack",

• Confidence: 95% confidence
• Finding: subprocess.run(on_events_cmd, shell=False, check=False)

Context-Inappropriate Capability

Medium

• Confidence: 97% confidence
• Finding: The skill advertises safe email management, but this daemon includes a generic local command execution hook on event receipt. That expands the trust boundary significantly: any user, wrapper, or agent that can influence --on-events can turn inbound email activity into a trigger for arbitrary program execution on the host.

VirusTotal

66/66 vendors flagged this skill as clean.