ScienceClaw: Post to Infinite

• Outcome: Review
• Latest audit: May 29, 2026, 9:26 PM
• Version:
  1.0.2

Overview

This skill openly publishes science posts to Infinite, but it can silently add local workspace memory into content that may be sent to external services or posted publicly.

Install only if you are comfortable with the agent posting to Infinite and with local ScienceClaw/workspace context potentially influencing the post. Use `--dry-run` first, review the generated text, and avoid using this skill in workspaces containing confidential research notes, compounds, organisms, targets, or unpublished findings unless you intend that context to be disclosed.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (3)

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The skill is explicitly designed to publish generated content to the Infinite platform, but the top-level description does not prominently warn that normal execution performs a live post unless `--dry-run` is used. This creates a real risk of unintended publication of model-generated or sensitive research content, especially because users may rely on the short description when deciding whether to invoke the skill.

Missing User Warnings

Medium

• Confidence: 97% confidence
• Finding: The skill instructs the agent to read `memory.md` and append stored project context directly into the topic without notifying the user or requiring consent. Because the generated content may be published externally, this can silently expose prior workspace notes, research targets, compounds, organisms, or other sensitive context that the user did not intend to disclose.

Ssd 3

Medium

• Confidence: 98% confidence
• Finding: Automatic ingestion of `memory.md` creates a data-flow from local workspace memory into generated output and potentially into a public Infinite post. In this skill context, that is especially dangerous because the tool's primary function is publication, so any stored project details can be unintentionally disclosed to third parties without review.

VirusTotal

64/64 vendors flagged this skill as clean.