Back to skill

Security audit

脚本转视频 Script to Video

Security checks across malware telemetry and agentic risk

Overview

The skill is a disclosed wrapper for the dLazy video-generation CLI, with normal SaaS risks around prompts, uploaded files, API keys, and npm-based installation.

Install only if you intend to use dLazy’s hosted service. Review the CLI source/package, prefer npx or a local install over a global install when possible, and only attach files you are comfortable uploading to dLazy. Treat the saved API key like any other service credential and rotate it if exposed.

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:4
Finding
Unverified Third-Party CLI Is Downloaded and Executed## Vulnerability Details **File Location**: `SKILL.md:4` and the duplicated configuration in `SKILL-cn.md:4` **Vulnerability Type**: Third-party dependency and software supply-chain exposure **Risk Level**: Medium **Complete Code Snippet**: ```text metadata: {"clawdbot":{"emoji":"🎞️","requires":{"bins":["npm","npx"]},"install":"npm install -g @dlazy/cli@1.2.3","installAlternative":"npx @dlazy/cli@1.2.3","homepage":"https://github.com/dlazy-ai/cli","source":"https://github.com/dlazy-ai/cli","author":"dlazyai","license":"see-repo","npm":"https://www.npmjs.com/package/@dlazy/cli","configLocation":"~/.dlazy/config.json","apiEndpoints":["api.dlazy.com","files.dlazy.com"]},"openclaw":{"systemPrompt":"When this skill is called, run 'dlazy chat --skill storyboard --prompt ...' for a new task, or 'dlazy chat --project <id> --prompt ...' to continue (discover ids via 'dlazy projects list'). Never pass both --skill and --project."}} ``` The associated installation instructions at `SKILL.md:48-55` and `SKILL-cn.md:48-55` explicitly recommend executing the dependency through `npx` or installing it globally. ### Technical Analysis The Skill delegates its functionality to the externally distributed npm package `@dlazy/cli`. Pinning the dependency to version `1.2.3` limits unexpected version changes, but the project provides neither a lockfile nor an integrity digest that binds installation to a specifically audited package artifact. Both `npx @dlazy/cli@1.2.3` and `npm install -g @dlazy/cli@1.2.3` retrieve package content from the npm distribution channel and execute package-controlled code. npm lifecycle scripts may also execute during installation unless explicitly disabled. Consequently, compromise of the maintainer account, package release, registry delivery path, or another component in the package's dependency tree could result in arbitrary code execution even though the Skill documents themselves contain no embedded malicious code. ...[truncated 1723 chars]
Remediation
## Remediation Suggestions 1. Audit the exact npm package archive and its complete transitive dependency tree before approving it for Agent execution. 2. Record and verify a trusted cryptographic integrity digest for the approved package artifact rather than relying only on a semantic version. 3. Use a lockfile or an equivalent immutable dependency manifest that includes integrity metadata for all transitive dependencies. 4. Prefer an isolated, project-local installation over `npm install -g` to limit persistence and environmental impact. 5. Disable npm lifecycle scripts with `--ignore-scripts` where compatible with the package, or explicitly review every required lifecycle script before allowing execution. 6. Run the CLI in a restricted sandbox with only the network, filesystem, and credential access required for the requested task. 7. Provide the dLazy credential only to the isolated process and prevent access to unrelated environment variables, SSH keys, browser data, and user files. 8. Require explicit user confirmation before uploading local files, particularly when those files may contain confidential or regulated information. 9. Continuously monitor the approved package version, maintainer ownership, release provenance, and dependency advisories for supply-chain changes.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (2)

Vague Triggers

Medium
Confidence
89% confidence
Finding
The trigger list includes broad terms such as 'storyboard' and 'character animation' that can appear in many normal creative requests, increasing the chance this skill is invoked when the user did not specifically ask to use an external SaaS-backed video-generation tool. Because the skill uploads attached files and sends prompts to remote endpoints, accidental activation can lead to unintended data disclosure and unnecessary external actions.

Vague Triggers

Medium
Confidence
93% confidence
Finding
The top-level description says to use the skill when the user gives a script or scene breakdown and wants a video, which is broad enough to match many ordinary requests about scripts, editing, or planning. In this skill's context, that is more dangerous because invocation routes user content to a third-party API and may upload local files, so an overbroad matcher can cause unintentional exfiltration or tool use without sufficiently explicit user intent.

VirusTotal

59/59 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.