Back to skill

Security audit

SystemTemp

Security checks for vulnerabilities and agentic risk

Overview

This temperature-monitoring skill is mostly local and purpose-aligned, but its installer can fetch unpinned remote Node code that is not clearly tied to the reviewed files.

Review the exact Node package source before installing, or prefer a pinned/local install whose temp executable is traceable to the reviewed temp.js. Expect the tool to create temperature logs and alert configuration under ~/openclaw_workspace, and avoid the /usr/local/bin copy method on shared or managed machines unless you intentionally want a system-wide command.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
Findings (1)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:13
Finding

Unpinned Remote Node.js Package Installation

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 13–18
Vulnerability Type: Unpinned third-party package installation
Risk Level: Medium

Vulnerable Code

yaml
{
  "id": "node",
  "kind": "node",
  "package": "SystemTemp",
  "bins": ["temp"],
  "label": "Install SystemTemp (Node.js)",
},

Technical Analysis

The installation metadata identifies the registry package SystemTemp without specifying an exact version, package integrity hash, or immutable trusted source. Consequently, the code retrieved during installation can change after this skill has been reviewed.

The local package.json also has no bin mapping for the advertised temp executable. This creates an additional provenance mismatch: the temp binary supplied by the remote package cannot be established from the audited local manifest as the reviewed temp.js implementation.

This is a software supply-chain weakness. It does not prove that the current registry package is malicious, but it permits a compromised publisher account, malicious future release, package replacement, or resolution of an unintended package to introduce code outside the audited artifact.

Attack Path

  1. An attacker compromises the publisher account or otherwise gains control of the referenced SystemTemp registry package.
  2. The attacker publishes a malicious release containing a lifecycle script or malicious temp executable.
  3. A user installs the skill using the declared Node.js installation metadata.
  4. Because no exact version or integrity value is specified, the installer resolves mutable registry content.
  5. Malicious lifecycle code may run during installation, or the malicious executable may run when the user invokes temp.

Impact Assessment

Successful exploitation would execute attacker-controlled code with the privileges of the account performing the installation or invoking the installed CLI. Depending on those privileges and the malicious payload, this could exp ...[truncated 254 chars]

Remediation
View remediation

Remediation Suggestions

  1. Pin the dependency to a reviewed exact version rather than using the mutable package name alone.
  2. Record and verify the package integrity hash through an appropriate lockfile or installer integrity field.
  3. Prefer installation directly from the reviewed local artifact or an immutable, verified source.
  4. Add an explicit bin mapping to the local package.json, for example mapping temp to temp.js, so the installed executable is traceable to the audited implementation.
  5. Verify package ownership, publisher identity, and repository provenance before distribution.
  6. Review and restrict package lifecycle scripts; disable them during installation where they are unnecessary.
  7. Re-audit the exact packaged tarball and compare it with the repository before publishing or installing it.
Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (12)

Tp4

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding

The documented behavior does not match the detected behavior: the skill claims temperature/CPU/fan monitoring, but reportedly performs undeclared local filesystem writes and does not actually implement some advertised monitoring features. This mismatch is dangerous because users may grant trust to a monitoring-only skill while it silently persists data/configuration locally, increasing the chance of unexpected data retention, tampering, or abuse of the local workspace.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The entire README, including user examples and command guidance, is written only in Traditional Chinese, and the examples imply the interaction language is fixed. There is no indication that users can choose another language or that the locale restriction is intentional for a region-specific use case.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The README explicitly advertises history, alerts, reports, and a data directory containing logs and configuration, but it does not clearly warn users that the skill persists operational data to disk. This can create privacy and operational risk because users may unknowingly retain telemetry and alert settings in predictable locations, especially on shared systems.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The installation instructions recommend copying executables into /usr/local/bin, which modifies a system-wide executable path and may require elevated privileges, but the README does not warn about those consequences. On multi-user or managed systems, this can lead to unintended persistence, path hijacking concerns, or unsafe use of sudo without informed consent.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
80% confidence
Finding

The skill exposes shell/environment-backed capabilities but does not declare any explicit tool scope or permissions, which weakens reviewability and least-privilege enforcement. In a skill that reads system hardware state and writes local logs/configs, undeclared execution capability can enable broader command or environment access than users would reasonably expect.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill silently creates persistent directories and files for logs and configuration under the user's home directory even though its stated purpose is simple temperature monitoring. This expands the skill's scope from read-only hardware inspection to filesystem state management, which can surprise users, leave residual data, and create opportunities for misuse or privacy issues if the stored data is later consumed by other components.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The script silently creates and modifies persistent files for logs and config during initialization without any user-facing disclosure or opt-in. Hidden writes are risky because users may believe the tool is passive/read-only, while it actually leaves artifacts and changes local state that may affect privacy, auditing, or trust.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The code logs historical temperature data and modifies markdown documentation content, which exceeds the declared monitoring function and performs ongoing stateful writes. Unnecessary persistent writes increase attack surface, can corrupt unrelated user data if paths are reused, and make the skill behave more like a local state manager than a simple sensor reader.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

Temperature logging and README updates append persistent records without notifying the user, which creates undisclosed data retention. Even if the data is low sensitivity, covert persistence violates least surprise and can accumulate environmental or operational history that other local users or tools may later access.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

Updating markdown/README-style content is not necessary for collecting or displaying hardware temperatures and represents out-of-scope modification of local content. This can overwrite or append to user-maintained files, hide activity in seemingly innocuous documentation, and makes the skill less transparent about what data it changes.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The package description and keywords are presented in Chinese, which indicates a language-specific skill presentation without any visible opt-in, alternative locale, or justification that the skill is region-specific. Under the policy rule, forcing a specific language without user choice is a natural-language policy concern.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

The file’s natural-language strings and locale formatting are fixed to Traditional Chinese, including toLocaleString('zh-TW') and all command help/output text. This forces a specific language/locale without opt-in, which matches the policy category for language or locale constraints.

Content

No source excerpt is available for this finding.