ClawHub
Back to skill

Security audit

TinkerClaw Backlink Audit

Security checks for vulnerabilities and agentic risk

Overview

The skill is a disclosed backlink-audit tool with opt-in credential, network, and state-writing behavior, and no evidence of hidden exfiltration or unsafe automatic execution.

Before installing, replace the shipped allowlist and build-history SERIES with your own data. Only run --source backlinks if you are comfortable sending the target domain and your backlinks.sh API key to api.backlinks.sh, and only run build-history.mjs --yes if you intend to write those exact historical metrics into your OpenClaw control-panel store.

A.I.G

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

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Description-Behavior Mismatch

Medium
Confidence
92% confidence
Finding
The script is not limited to backlink discovery/classification; it is explicitly designed to write historical backlink metrics into an OpenClaw control-panel store. Even though it requires a --yes flag and labels the data as the author's example dataset, this still introduces a state-mutating capability that exceeds the skill's described audit/report scope and could pollute telemetry or dashboards with misleading data if run by a user who does not fully understand it.

Context-Inappropriate Capability

Medium
Confidence
90% confidence
Finding
The script invokes an external CLI to perform control-panel.record writes, giving the skill a side-effecting capability unrelated to simple backlink auditing. Because the command mutates persistent control-panel state, a user running the skill could unknowingly alter another component's data store, and the skill context does not justify that privilege as part of normal backlink analysis.

Static analysis

Detected: suspicious.env_credential_access

Environment variable access combined with network send.

Critical
Code
suspicious.env_credential_access
Location
scripts/backlink-audit.mjs:153