Back to skill

Security audit

centsible

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill benign with medium confidence: scanners found configuration changes and software updates beyond its stated purpose, but no malicious behavior was reported; Skillspector’s review was partial.

The verdict is benign, but the stated purpose does not clearly cover all reported settings changes and software updates. Consider whether you want a skill that can make those changes; the evidence is incomplete because Skillspector marked its analysis partial after a local path-like reference could not be resolved.

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
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The supplied commands perform software maintenance rather than repository fact management or token optimization inspection. Updating installations is a distinct, undeclared capability that modifies software outside the stated fact-management purpose.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The commands change resource-budget settings; they do not manage repository facts, record discoveries, or inspect optimization status. Budget modification is a materially distinct capability not disclosed by the description. The effect of unsetting the token budget depends on Centsible's defaults, but the configuration change itself is explicit.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
91% confidence
Finding

The declared purpose covers managing reusable knowledge and inspecting optimization status. The supplied commands instead primarily alter tool configuration, including workflow controls such as reflection hooks and error limits. Configuration listing aligns with inspection, and optimization settings are related to token efficiency, but the broader configuration-writing capabilities are not disclosed. No micro-fact management or discovery recording is shown.

Content

No source excerpt is available for this finding.