ClawHub
Back to skill

Security audit

Quasi Coder

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed coding helper for turning shorthand into scoped code edits, with some broad but visible guidance about running requested supporting actions.

Install this if you want an agent to convert rough shorthand into code, but keep requests explicit: name the target file, mark the region, and review diffs before accepting changes. Treat any command execution, data fetching, or file creation as something you should request deliberately and keep narrowly scoped.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (6)

Anti-Refusal Statement

High
Category
Anti-Refusal
Content
1. **Focus on Core Mechanisms**: Implement the essential functionality that makes the project work
2. **Apply Expert Knowledge**: Use computer science principles, design patterns, and industry best practices
3. **Handle Imperfections Gracefully**: Work with typos, incorrect terminology, and incomplete descriptions without judgment
4. **Consider Context**: Look at available resources, existing code patterns, and project structure
5. **Balance Vision with Excellence**: Respect the collaborator's vision, explicit constraints, and file boundaries while ensuring technical quality
6. **Avoid Over-Engineering**: Implement what's needed, not what might be needed
Confidence
80% confidence
Finding
Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.

Vague Triggers

Medium
Confidence
90% confidence
Finding
The trigger design includes broad, generic activation terms and generalized shorthand conventions, making it likely the skill activates on ordinary requests rather than clearly scoped invocations. Ambiguous activation raises the risk that the skill's expansive behavior is applied unexpectedly, including code changes or side-effectful actions in the wrong context.

Context-Inappropriate Capability

Medium
Confidence
95% confidence
Finding
The skill explicitly permits non-code actions such as running commands, creating files, and fetching data, which expands it from code translation into operational behavior. That broader authority can be abused to perform unintended side effects, exfiltrate data, or alter the environment based on ambiguous shorthand instructions.

Context-Inappropriate Capability

Medium
Confidence
96% confidence
Finding
The troubleshooting section instructs the skill to execute non-code actions on request, including commands and data fetching, despite the skill being presented as a coding assistant. This weakens safety boundaries by normalizing operational execution from natural-language shorthand, increasing the chance of harmful or unauthorized actions.

Vague Triggers

Medium
Confidence
95% confidence
Finding
Allowing the skill to activate when shorthand intent is merely 'clear' removes a hard boundary and makes invocation subjective. Subjective triggering is dangerous because an attacker or mistaken user can smuggle instructions into ordinary text and cause the skill to perform unintended transformations or side effects.

Context-Inappropriate Capability

Medium
Confidence
94% confidence
Finding
The advanced example endorses fetching external API data and saving it to a file, which introduces network access and filesystem side effects beyond the stated role of interpreting shorthand into code. Even as an example, it teaches and legitimizes behavior that could be triggered to access external resources or write unreviewed artifacts.

Static analysis

No suspicious patterns detected.