Back to skill

Security audit

AgentData

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill benign with high confidence: the completed scans found no AIG issues, while Skillspector flagged two low-probability agency patterns that its review judged consistent with the instructions shown.

The verdict is benign, not risk-free. Before installing, consider whether you are comfortable configuring a dedicated wallet and private key for paid calls, and review the payment caps and per-call approval behavior described in the scan excerpts.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 63)May include surrounding context.

md
The npm package `agentdata-mcp` pays automatically from a buyer wallet, with a per-call cap
(`AGENTDATA_MAX_PAYMENT_USDC`, default 0.05) and a per-process budget
(`AGENTDATA_SESSION_BUDGET_USDC`, default 1.00). It exposes every endpoint as a tool.
The four free tools are marked read-only, so OpenClaw can run them without asking; the paid ones are
not, so it asks before each call that would pay. Leave it that way.

The user sets this up, not you:

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
80% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 70)May include surrounding context.

md
1. A **dedicated** wallet holding only a few USDC on Base. Never a main wallet.
2. Its private key in the environment of the process that runs OpenClaw, for example
   `export AGENTDATA_BUYER_PRIVATE_KEY=0x...` before starting the gateway. Never ask the user to
   paste a key into the chat, and never print it.
3. Then register the server. The key is referenced, not written into the config; OpenClaw fills
   `${...}` in from its own environment when it starts the server: