ClawHub
Back to skill

Security audit

Work Productivity Agent Browser Workflow Helper

Security checks for vulnerabilities and agentic risk

Overview

This is a documentation-only workflow helper with overly broad activation wording, but I found no hidden execution, data access, persistence, or destructive behavior.

Install only if you want a broad workflow helper for Agent Browser-style planning and reliability work. Be aware that its generic trigger terms may activate it for some unrelated browser, CLI, or bug-fix requests, so confirm the selected skill when precision matters.

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 (10)

Vague Triggers

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

The trigger sentences are broad, templated, and partially match generic user requests like asking for help, needing a workflow, or invoking browser/automation-related work. This increases the chance of unintended skill activation, causing the agent to apply this workflow in contexts the user did not explicitly intend, which can misroute tasks or lead to unsafe automation being used without clear consent.

Content

No source excerpt is available for this finding.

Vague Triggers

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The trigger phrases are broad, repetitive, and partially generic, which increases the chance the skill will be invoked when a user did not specifically intend to select it. In an agent/browser automation context, unintended invocation can cause the system to apply the wrong workflow, generate misleading actions, or steer execution toward browser/headless automation tasks that carry higher operational risk than ordinary text assistance.

Content

No source excerpt is available for this finding.

Vague Triggers

High
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest description includes very broad activation terms such as 'browser', 'automation', and 'practical workflow' that can match many unrelated user requests. This increases the chance the skill is invoked outside its intended scope, causing prompt hijacking of benign tasks, incorrect delegation, or unintended exposure of user context to an irrelevant skill.

Content

No source excerpt is available for this finding.

Vague Triggers

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The trigger keywords are overly generic and ambiguous, including terms like 'browser', 'headless', 'automation', and 'cli', which are common across many unrelated tasks. In a skill-routing system this can cause over-selection, letting the skill intercept prompts it should not handle and amplifying the risk of misrouting, unsafe actions, or contamination of downstream agent behavior.

Content

No source excerpt is available for this finding.

Vague Triggers

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The trigger examples include a very broad natural-language phrase ('Help me ...') tied to generic topics like browser, automation, and workflow support. In agent systems that auto-route based on fuzzy matching, this can cause unintended invocation, leading the wrong skill to run on unrelated requests and potentially produce unsafe or confusing browser-automation guidance.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
83% confidence
Finding

The file is a zh-CN README, but all listed trigger phrases are English-only and there is no indication that users may invoke the skill in Chinese or choose their preferred language. This can amount to a language/locale policy issue because the skill appears to force a specific language without user opt-in or justification.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The invocation description says to use the skill whenever users mention broad categories like work-productivity, browser, headless, or automation, which are common across many benign and unrelated tasks. This ambiguity makes unintended invocation likely and can cause the agent to apply this skill in contexts where its assumptions, workflows, or generated artifacts are a poor fit, reducing safety and predictability.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The trigger keyword list includes very generic terms such as "browser", "automation", "cli", and "bug fix", which can match many unrelated requests and cause the skill to activate outside its intended scope. Over-broad invocation can route user tasks into the wrong workflow, increasing the chance of inappropriate guidance, unnecessary tool use, or accidental execution of browser-automation-oriented steps in contexts that did not request them.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The default_prompt uses a very broad auto-invocation phrase ('Use $work-productivity-agent-browser-workflow-helper to help me ...') tied to expansive terms like work-productivity, browser, headless, automation, and implementation support. This can cause the skill to be invoked in situations outside its intended scope, increasing the chance of unsafe or irrelevant automation guidance being injected into unrelated conversations.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The invocation guidance is ambiguous and malformed, with truncated trigger sentences and broad keywords that overlap many ordinary user requests. This increases the chance of accidental skill selection or misrouting, which is especially risky here because the skill can influence browser automation, setup, and safety decisions users may trust without noticing the mismatch.

Content

No source excerpt is available for this finding.