Back to skill

Security audit

Ai Agent Setup

Security checks for vulnerabilities and agentic risk

Overview

This is a short consulting-style AI agent setup skill with no bundled code or hidden behavior, though users should treat deployment advice and the unpinned install command cautiously.

Before installing, verify the ClawHub package source and consider using a pinned reviewed version instead of @latest. Treat any real customer-platform deployment as a separate action requiring explicit approval, scoped credentials, and environment checks.

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:12
Finding
Unpinned Third-Party Package Execution Through npx## Vulnerability Details **File Location**: `SKILL.md`, lines 12-14 **Vulnerability Type**: Unpinned executable dependency **Risk Level**: Medium ```bash npx clawhub@latest install ai-agent-setup ``` ### Technical Analysis The installation instructions invoke `clawhub` through `npx` using the mutable `latest` tag. If the package is not already available locally, `npx` may download and execute the version currently published under that tag. The project does not specify an audited version, integrity hash, lockfile, or trusted package provenance. Because `latest` can point to different package contents after this skill has been reviewed, the command creates a supply-chain risk. Compromise of the package publisher, registry account, release process, or package distribution channel could cause users to execute altered code without any corresponding change to this repository. ### Attack Path 1. An attacker compromises the `clawhub` package publisher, release pipeline, registry account, or another relevant distribution component. 2. The attacker publishes a malicious release and assigns it to the `latest` tag. 3. A user follows the documented installation command. 4. `npx` retrieves and executes the attacker-controlled package version. 5. The malicious package executes with the permissions of the user running the command and may modify files, access user-readable information, invoke network services, or install additional components. ### Impact Assessment Successful exploitation can provide arbitrary code execution under the invoking user's account. The accessible scope includes files, credentials, environment variables, network resources, and applications available to that account. If the command is run from a privileged administrative or CI/CD context, the impact may expand to system-wide resources, deployment credentials, build artifacts, or connected infrastructure. The reviewed file does not itself demonstrate privilege es ...[truncated 66 chars]
Remediation
## Remediation Suggestions - Replace the mutable `latest` tag with an exact, reviewed package version, for example `npx clawhub@X.Y.Z`. - Verify that the package comes from the intended registry and publisher before execution. - Use registry-supported integrity verification or a documented cryptographic checksum where available. - Review the pinned package, including lifecycle scripts and transitive dependencies, before recommending it. - Execute installation with a non-privileged account in an isolated environment. - In automated environments, enforce dependency allowlists, lockfiles, and reproducible installation controls. - Establish a controlled process for reviewing and explicitly approving upgrades rather than tracking `latest` automatically.
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill explicitly mentions deploying agents to customer platforms but provides no warning, guardrail, or confirmation requirement for system-impacting actions. In context, this increases risk because deployment can affect real customer environments, credentials, integrations, and production workflows if the skill is invoked inappropriately or followed too literally.

Rp1

Medium
Category
MCP Rug Pull
Confidence
95% confidence
Finding
The installation command uses `npx clawhub@latest`, which pulls the latest published package rather than a pinned, reviewed version. This creates a supply-chain risk: a compromised or malicious future release could be executed by users during installation with no version stability or trust boundary.

Vague Triggers

Medium
Confidence
93% confidence
Finding
The trigger condition "當用戶話『整AI Agent』既時候" is broad and can overlap with ordinary conversation about building agents, causing unintended activation. In a skill that discusses consulting, design, quoting, and deployment, accidental invocation could lead the agent to take or propose actions outside the user's precise intent.

Static analysis

No suspicious patterns detected.