Back to skill

Security audit

0-sentisense-onboarding

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marks this skill suspicious because it reports unpinned `npx sentisense` commands that could run a future package release locally, plus a quote lookup capability outside the stated onboarding purpose.

Review the skill before installing. Consider whether you need it to run a CLI at all; the scanner notes an HTTPS provisioning option. If you use the CLI, prefer a reviewed exact version and restrict what it can access. The scan is incomplete: a local path-like reference was unresolved, so conclusions may not cover everything the skill references.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The supplied chunk performs a concrete market-data operation. It neither configures an API key nor provides task-routing guidance. Although using an existing API key is consistent with the broader SentiSense context, retrieving an AAPL quote is a distinct capability not represented in the declared purpose.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx sentisense'.

Medium
Category
MCP Rug Pull
Confidence
96% confidence
Finding

The recommended npx sentisense init command can download and execute an unpinned npm release with the agent's local permissions. Although this is a CLI rather than an MCP server, the underlying supply-chain risk is real: a compromised future release could access local files or credentials. The documented HTTPS provisioning alternative reduces the need for this exposure.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx sentisense'.

Medium
Category
MCP Rug Pull
Confidence
96% confidence
Finding

This executable example repeats the unpinned npm command, allowing the package release selected at execution time to run locally. A compromised release could execute arbitrary code with the agent's permissions; there is no evidence here that the current package is malicious. This is another occurrence of the same supply-chain issue, not a separate attack path.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx sentisense'.

Medium
Category
MCP Rug Pull
Confidence
92% confidence
Finding

This paragraph references the same unpinned execution route rather than introducing another executable example. Its disclosure of shell, registry, and configuration access makes the behavior transparent, but does not constrain which package version runs or what that code can access. It therefore inherits the earlier supply-chain vulnerability rather than constituting an independent vulnerability.

Content

No source excerpt is available for this finding.