ClawHub
Back to skill

Security audit

CSPR Memory Curator

Security checks for vulnerabilities and agentic risk

Overview

The skill has a coherent preference-memory purpose, but it can update persistent user profile memory on a broad trigger without explicit confirmation or rollback guidance.

Install only if you want an agent to curate cspr preference memory from prior feedback and recent outputs. Before use, ask the agent to show the proposed profile diff and get your confirmation before running `prefmem update`, especially because the resulting profile can influence future sessions.

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 (1)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill says to use it "after the user has provided feedback or after several runs," which is a broad trigger that can cause the agent to access and rewrite persistent preference memory without a tightly scoped authorization condition. Because the workflow reads from and writes to user state files and applies the result with a command, loose activation criteria increase the chance of unintended or premature memory modification from ambiguous context.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.