Back to skill

Security audit

skill-find

Security checks for vulnerabilities and agentic risk

Overview

The skill is not overtly malicious, but it recommends unpinned command-line package execution and silent global skill installation that users should review carefully.

Install only if you are comfortable reviewing third-party skills and commands first. Avoid running the shown `npx` commands as-is in a sensitive environment; prefer pinned, trusted CLIs, remove `-y`, avoid global installs unless needed, and inspect the target skill before installation.

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:36
Finding

Unpinned Third-Party Package Execution and Non-Interactive Global Skill Installation

Content
View full analysis
npx skills add -g -y ``` `SKILL.md`, line 75: ```bash npx skills init [new-skill-name] ``` ### Technical Analysis The documented workflow instructs the agent to invoke third-party command-line packages through `npx` without pinning exact package versions or verifying package integrity. If a requested package is not already available locally, `npx` can retrieve and execute it from the configured package registry. The installation workflow also recommends `npx skills add` with both the global installation flag (`-g`) and automatic confirmation flag (`-y`). This combination can install third-party skill content persistently in the user's environment without an interactive confirmation step. The instructions do not require cryptographic signature verification, checksum validation, publisher allowlisting, package-lock validation, or review of lifecycle scripts. The document advises users to read the downloaded `SKILL.md`, verify prerequisites, and test in isolation. These are useful procedural precautions, but they do not technically prevent execution of a compromised CLI package, malicious lifecycle script, dependency-confusion package, typosquatted package, or compromised skill repository. ### Attack Path 1. An attacker publishes a typosquatted package or skill, compromises a legitimate package or repository, or injects a malicious transitive dependency. 2. The malicious item is returned by a registry search or is supplied as the requested package or skill identifier. 3. The agent or user executes an unpinned `npx` command, causing the cur ...[truncated 1232 chars]
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (11)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 62)May include surrounding context.

md
1. 阅读 `SKILL.md` 的说明规范。

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 97)May include surrounding context.

md
1. 阅读 `SKILL.md` 的说明规范。

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The manifest description is broad enough to trigger the skill on many generic requests about finding functionality or discovering skills, which can cause over-invocation of a skill that recommends executing external discovery and installation commands. In context, broad activation increases exposure to the risky supply-chain workflows described elsewhere in the file.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The file mixes an English heading with predominantly Chinese operational instructions, which effectively imposes a language preference in the skill's natural-language interface. Under the policy, locale or language constraints should be opt-in or clearly justified; neither is present here.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
92% confidence
Finding

The skill instructs users to run npx skills without pinning an exact package version, which allows execution of whatever package version is current at invocation time. If the package is compromised, typo-squatted, or updated maliciously upstream, users may execute attacker-controlled code on their machine during search or installation workflows.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
95% confidence
Finding

The command npx clawhub search executes an unpinned remote npm package, creating a supply-chain risk because package contents can change over time without review. In a skill whose purpose is to discover and then install more third-party skills, this compounds trust risk by normalizing execution of external code before any validation.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
92% confidence
Finding

This use of npx skills find is another unpinned transient package execution path and exposes users to upstream package compromise or unexpected behavior changes. Because the skill positions the command as part of routine discovery, users may run it with minimal scrutiny, increasing exploitation likelihood.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
94% confidence
Finding

The sorting example again uses npx clawhub without a pinned version, leaving execution dependent on the current upstream npm state. This is dangerous in context because the skill is a discovery gateway to further third-party code, so compromise here can steer users toward malicious results or run arbitrary code immediately.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding

npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
97% confidence
Finding

npx skills add <owner/repo@skill> -g -y combines unpinned remote package execution with silent global installation, reducing user review and increasing blast radius if the package or referenced skill is malicious. The -y and global install flags make the workflow more dangerous because they encourage unattended acceptance and system-wide changes.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
90% confidence
Finding

Recommending npx skills init without version pinning still exposes users to execution of the latest upstream package code. Although this path is for skill creation rather than installation, it remains a supply-chain execution risk if the npm package is compromised.

Content

No source excerpt is available for this finding.