ClawHub
Back to skill

Security audit

sn-prepare-citations

Security checks across malware telemetry and agentic risk

Overview

This skill locally reformats research report citations and writes report outputs; users should be careful because legacy mode can overwrite the input report.

Install only if you want a local helper that rewrites markdown research reports from explicit input files. Prefer the documented --output workflow so the original stitched.md remains intact, and review generated report.md and citations.json before publishing or sharing them.

SkillSpector

Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
91% confidence
Finding
The skill documentation describes capabilities that read input files, write output files, and process evidence-derived source metadata, yet no permissions are declared. That creates a trust and enforcement gap: a caller may invoke the skill believing it is low-risk text formatting, while it can access and modify local artifacts and potentially fetch network-referenced content depending on implementation. In an agent environment, undeclared capabilities weaken policy controls and reviewability.

Tp4

High
Category
MCP Tool Poisoning
Confidence
89% confidence
Finding
The declared purpose says the skill only post-processes citations, but the documented behavior also modifies document structure (L0/TOC), repairs claim-id leakage, and emits an additional citations.json artifact. This mismatch can cause downstream systems or reviewers to approve the skill under a narrower risk model than its actual behavior, increasing the chance of unintended content changes or metadata leakage handling outside expected controls.

Missing User Warnings

Medium
Confidence
87% confidence
Finding
The legacy compatibility mode overwrites the input report file in place when --outline/--output are omitted, but the documentation does not prominently warn about destructive behavior. In automated pipelines, this can lead to irreversible data loss, corruption of intermediate artifacts, or accidental modification of the only copy of a report if the mode is invoked incorrectly.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.