Back to skill

Security audit

tgtrust

Security checks for vulnerabilities and agentic risk

Overview

The verdict is benign because no AIG findings were reported and the main concerns are that the skill’s described verification purpose does not match reported key-generation and registration steps; the scan was partial.

The verdict is benign, but consider whether you intend to generate a local key and register a credential with an issuer, rather than only verify another agent. Skillspector’s analysis was partial because two referenced local paths were missing, and the registration command’s internal effects could not be determined.

Vulnerability Patterns
  • 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
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The supplied command runs tgtrust.client register, not a credential or stake verification operation. Registration is materially different from the declared purpose of checking an unknown counterparty. The snippet shows no verification of another agent’s credential or TON stake; the registration command’s internal effects cannot be determined from this snippet alone.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The supplied command runs tgtrust.client new-key --out agent.key. It does not accept another agent's public key or perform a credential or stake check. Although key generation could support a broader tgtrust workflow, this chunk's sole operation is materially different from the declared verification purpose.

Content

No source excerpt is available for this finding.

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 84)May include surrounding context.

saves the credential file. Then:

bash
python3 -m tgtrust.client register --credential credential.json --key-file agent.key

Keep agent.key private; never print it or send it to anyone.

The skill expands from checking another agent's credential to creating and registering its own identity.

Medium
Category
Excessive Agency
Confidence
91% confidence
Finding

The manifest at L03 describes checking another agent's credential and optional stake. The workflow at L74–L85 instead instructs the agent to generate a private-key file and register its own credential with an issuer, adding local and remote state-changing operations beyond the declared verification purpose.

Content

No source excerpt is available for this finding.