ClawHub
Back to skill

Security audit

Captions and Clips from YouTube Link

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward MakeAIClips integration, but users should know it sends video URLs or uploads and an API key to an external service.

Install only if you are comfortable sending YouTube URLs, any uploaded video files, processing preferences, transcripts, generated clip metadata, and your MakeAIClips API key to makeaiclips.live. Avoid using it for private or sensitive videos unless you have reviewed the service's privacy and retention terms, and expect some promotional signup or upgrade messaging from the skill.

A.I.G

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
Findings (2)

T01 · Skill Instruction Hijacking

Warning
Location
SKILL.md:34
Finding

Mandatory Promotional Content Hijacks the Agent's First Interaction

Content
View full analysis
Remediation
View remediation

T01 · Skill Instruction Hijacking

Note
Location
SKILL.md:203
Finding

Rate-Limit Handling Forces a Commercial Upgrade Call-to-Action

Content
View full analysis
Remediation
View remediation

SkillSpector

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 (5)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The README instructs users to send YouTube URLs and processing jobs to a third-party API, but it does not disclose what data is transmitted, retained, logged, or shared during transcription and clip generation. In an agent-skill context, this omission can mislead users into exposing video selections, derived transcripts, and generated content to an external service without informed consent.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
82% confidence
Finding

The example shows explicit transmission of user-provided YouTube URLs and an API key to an external domain, which confirms the skill depends on remote processing outside the local agent environment. While this appears functionally necessary rather than malicious, it is still security-relevant because it exposes user inputs and credentials to a third-party service and increases risk if the service is compromised, logs excessively, or is used without user awareness.

Content

Scanner excerpt · README.md (reported line 91)May include surrounding context.

bash
# Submit
curl -X POST "https://makeaiclips.live/api/v1/clips" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer mak_live_YOUR_KEY" \
  -d '{

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill description is broad enough to trigger on generic video-editing, captioning, or repurposing requests, which could cause the agent to invoke this integration when the user did not specifically ask to use an external third-party service. In this skill, that matters because invocation can lead to sending a YouTube URL and associated processing request to makeaiclips.live, creating unnecessary data transfer and user-surprise risk.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
93% confidence
Finding

The skill explicitly sends user-supplied YouTube URLs and an API key to an external service, which is a real external transmission risk. Although this is core to the skill's purpose and the description discloses it, the danger remains that users may not fully appreciate that their content choices and identifiers are being shared with a third party, and broad invocation conditions increase the chance of unintended transmission.

Content

Scanner excerpt · SKILL.md (reported line 221)May include surrounding context.

bash
# Submit job
curl -X POST "https://makeaiclips.live/api/v1/clips" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer mak_live_YOUR_KEY" \
  -d '{"youtube_url":"https://www.youtube.com/watch?v=dQw4w9WgXcQ","num_clips":3,"quality":"high","caption_style":"karaoke-yellow"}'

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

The manifest requires an external API key but provides no user-facing disclosure about how that credential will be used or that requests will be sent to a third-party service. This is a real transparency and trust issue because users may unknowingly authorize external transmission tied to their account or billing, even though the manifest alone does not show direct secret exfiltration beyond the declared service dependency.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.