ClawHub
Back to skill

Security audit

call-chain

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed code-analysis helper that traces call chains and does not show hidden persistence, exfiltration, or destructive behavior.

Install this if you are comfortable with the agent checking your local Claude plugin directory for the gauntlet helper and running that helper for code-graph queries. If gauntlet is not installed, the skill falls back to workspace code search.

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 (2)

Agent Config Directory Access

High
Category
Agent Snooping
Confidence
94% confidence
Finding

The skill instructs the agent to search the user's ~/.claude/plugins directory, which accesses agent configuration/plugin installation data outside the target project. Even though the stated purpose is plugin discovery, reading from a user-specific config area can expose sensitive local environment details and violates the principle of limiting skills to project-scoped resources. In this context, the danger is increased because the command explicitly enumerates files in a private agent directory rather than asking the user for the path or using a constrained integration API.

Content

Scanner excerpt · SKILL.md (reported line 27)May include surrounding context.

data. Discover it:

bash
GRAPH_QUERY=$(find ~/.claude/plugins -name "graph_query.py" -path "*/gauntlet/*" 2>/dev/null | head -1)

If gauntlet is not installed: Fall back to static

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest defines the trigger as "understanding how a function propagates through the system," which is a natural-language description broad enough to overlap with many ordinary code-analysis requests. The file does not provide narrower trigger phrases, exclusion conditions, or negative examples to clarify when this skill should or should not activate.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.