Content Repurposing

• Outcome: Review
• Latest audit: May 28, 2026 at 10:02 AM UTC
• Version:
  0.1.5

Overview

The skill is mostly a coherent content-repurposing helper, but it includes broad connected-service commands that can publish to a real X/Twitter account without an explicit confirmation safeguard.

Install only if you trust inference.sh and want this skill to use its CLI. Review every `infsh` command before running it, especially commands that upload private material or call posting apps such as `x/post-create`; treat generated social posts as drafts unless you explicitly approve publication.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands

Findings (4)

Description-Behavior Mismatch

Medium

• Confidence: 88% confidence
• Finding: The skill’s stated purpose is content repurposing, but it also includes an example that directly publishes to X/Twitter. That expands the capability from content transformation into external account action, which is security-relevant because generated or unreviewed content could be posted to a real account without sufficient user awareness or approval safeguards.

Context-Inappropriate Capability

Medium

• Confidence: 90% confidence
• Finding: Including direct social-media publishing is broader than the declared repurposing use case and violates least privilege for a formatting/adaptation skill. In practice, this can cause accidental or unauthorized posting if an agent invokes the example flow as part of a content transformation request.

Vague Triggers

Medium

• Confidence: 80% confidence
• Finding: The trigger list is broad and ambiguous, increasing the chance the skill is auto-selected for loosely related user requests. In an agent environment, over-broad matching can route benign content tasks into a skill that also contains account-affecting and external-tool instructions, raising the risk of unintended actions.

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The skill demonstrates posting to X/Twitter but does not warn the user that this action affects an external account and may create irreversible public output. Lack of a warning or confirmation step makes accidental publication more likely, especially when paired with autogenerated content.

VirusTotal

VirusTotal findings are pending for this skill version.