ClawHub
Back to skill

Security audit

Socrates

Security checks for vulnerabilities and agentic risk

Overview

This skill is a planning aid that asks the agent to briefly pressure-test its approach before complex work, with no code execution, data access, persistence, or hidden behavior found.

Install this if you want the agent to add a compact planning check before complex work. Be aware it may activate automatically for many planning, coding, refactoring, or delegation tasks, which can add extra structure or slow very small tasks unless your environment lets you disable or narrow implicit invocation.

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
Confidence
93% confidence
Finding
The manifest description says to use the skill when an agent is about to "plan, choose an approach, do non-trivial coding, refactor, decompose work, or delegate," which covers a very wide range of common agent behavior. Although the file later provides some skip/compress guidance, the primary activation text remains broad enough to risk unintended invocation across many routine tasks.

Vague Triggers

Medium
Confidence
92% confidence
Finding
The default prompt is broadly phrased around common activities like planning, coding, and delegation, which are routine across many agent tasks. This can cause the skill to activate too often, injecting extra instructions or reasoning steps into unrelated workflows and increasing the chance of prompt-surface expansion, unintended behavior changes, or instruction interference.

Vague Triggers

Medium
Confidence
96% confidence
Finding
Enabling implicit invocation without clear guardrails allows the skill to run automatically based on loosely matched contexts rather than deliberate user or system intent. In a skill designed to influence planning and approach selection, this broad automatic activation can unnecessarily affect many downstream decisions, create instruction collisions, and make agent behavior less predictable.

Static analysis

No suspicious patterns detected.