ClawHub
Back to skill

Security audit

Superpowers Dev Workflow

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed software-development workflow skill; its command use, subagents, commits, and branch actions fit that purpose and include user gates for higher-impact steps.

Install this if you want a structured, sometimes heavyweight coding workflow. Expect it to read project context, run test and git commands, create planning documents, make commits, and optionally use subagents; review branch merge, push, PR, or discard prompts carefully before approving them.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The top-level description includes broad trigger phrases such as "let's build," "help me plan," and "this is broken," which can match common conversational requests that are not necessarily meant to invoke a heavyweight code-execution workflow. In a skill that requires exec and sessions_spawn, accidental activation can lead to unnecessary tool use, expanded autonomy, and execution of a complex development pipeline when the user intended a simpler or safer interaction.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The Phase 1 trigger, "User wants to build something," is underspecified and does not constrain whether the request is about software, documentation, architecture discussion, or even non-code work. Because this skill enforces a mandatory workflow and later uses exec-enabled, subagent-driven steps, ambiguous activation increases the chance of over-privileged behavior being applied outside its intended scope.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The debugging trigger activates on "any technical issue," which is overly broad and can encompass tasks far beyond code debugging, including infrastructure, account, operational, or ambiguous troubleshooting requests. In the context of a skill designed to drive systematic debugging and potentially invoke execution tooling, this broad scope can cause the agent to adopt an invasive workflow and tool usage pattern without sufficient evidence that such escalation is appropriate.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.