Security audit

hypertaks

Security checks across malware telemetry and agentic risk

Overview

Hypertaks is a broad but disclosed orchestration skill that emphasizes user approval, scoped permissions, and safe handling before using tools or making changes.

Install this only if you want a broad CEO/founder-style orchestration layer that may ask intake questions and use multiple specialist perspectives. Review its contract before approving work, and require contract-ID approval for any file changes, external actions, memory setup, publishing, spending, deletion, deployment, or on-chain activity.

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 (5)

Vague Triggers

High

Confidence: 93% confidence
Finding: The trigger description is extremely broad, covering generic requests like analysis, building pages, coding, growth, and strategy. This creates a prompt-scope hijack risk where the skill activates for a large fraction of normal work, allowing its rigid protocol and authority model to dominate unrelated tasks and potentially interfere with safer or more specific skills.

Vague Triggers

High

Confidence: 90% confidence
Finding: The activation guidance says the skill applies whenever work is 'founder-shaped' and even mandates activation whenever the Boss names Hypertaks, but 'founder-shaped' is subjective and expansive. That ambiguity can cause over-activation, displacing task-specific safeguards and steering many ordinary requests into a powerful multi-agent workflow with broad domain claims.

Vague Triggers

Medium

Confidence: 94% confidence
Finding: The skill advertises very broad trigger conditions such as using Hypertaks whenever the Boss names it or needs 'founder-shaped work' across many domains. This can cause accidental activation on loosely related tasks, expanding the skill's authority surface and increasing the chance that heavyweight orchestration or file-generation behavior is invoked when a narrower skill would be safer.

Vague Triggers

Medium

Confidence: 96% confidence
Finding: The phrase 'founder-shaped work' is inherently ambiguous and the surrounding section instructs the skill to self-activate whenever that standard seems to apply. In a security context, ambiguous activation rules are dangerous because they let the agent reinterpret many requests as in-scope, potentially bypassing least-privilege expectations and causing overbroad use of tools, subagents, or output generation.

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The protocol treats generic affirmations like "approved", "go", or "yes" as sufficient to activate advisory read-only work. In a long conversational workflow, those tokens can easily appear as ordinary acknowledgments to a question, summary, or partial plan, creating ambiguity about whether the user intended to authorize substantive analysis under the contract. Because this skill is an orchestration/governance layer for broad business, engineering, finance, and potentially security-sensitive work, ambiguous activation increases the risk of unintended processing or analysis proceeding under mistaken consent.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.