ClawHub
Back to skill

Security audit

Code Share

Security checks for vulnerabilities and agentic risk

Overview

This skill shares code through GitHub Gist in a disclosed, purpose-aligned way, with the main risk being that selected files are uploaded to GitHub.

Install this only if you want the agent to share code through your GitHub account. Review files before sharing, keep the default secret setting unless public posting is intentional, and avoid using it for proprietary code or files that may contain credentials or personal data.

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

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

The declared description suggests a general code-sharing skill that posts code to GitHub Gist instead of inline chat, which implies creating or sharing a gist as the primary function. The actual code specifically edits an existing gist revision via gh gist edit and requires a gist ID/URL as input. It does not create a new gist, does not use the provided visibility argument, and depends on GitHub CLI/API access. While updating a gist is related to code sharing, the primary behavior is narrower and materially different from the declared purpose.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The script transmits the contents of a local file to GitHub Gist via gh gist create, but it provides no explicit warning, confirmation, or disclosure at the point of upload. In an agent-skill context, this is risky because users may not realize that potentially sensitive code, secrets, or proprietary data will leave the local environment and be stored on a third-party service, especially when the visibility can be set to public.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The script transmits a caller-supplied local file to GitHub Gist via gh gist edit without any explicit warning, confirmation, or safeguard about network exfiltration. In a code-sharing skill, this behavior is expected, but it still creates real risk because users may unintentionally upload sensitive local files, secrets, or proprietary code to a remote service, especially if they do not realize the action is networked and potentially public depending on gist settings.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.