Security audit

CSPR Memory Curator

Security checks for vulnerabilities and agentic risk

Overview

This skill is a small, disclosed preference-memory updater for CSPR, with no executable payload or hidden network behavior found.

Install this only if you want an agent to maintain your CSPR preference profile from accumulated feedback. Review generated profile updates before applying them if source hiding, disliked patterns, or long-term topic preferences could affect what news or sources you see.

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)

Vague Triggers

Medium

Confidence: 92% confidence
Finding: The invocation guidance 'Use this skill after the user has provided feedback or after several runs' is broad and underspecified, which can cause the agent to run a state-mutating skill without clear user consent or necessity. Because the skill reads feedback history and rewrites persistent preference memory, ambiguous triggering increases the chance of unintended profile changes and privacy-impacting state updates.

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The skill instructs the agent to read feedback history and the current profile, then write and apply an updated persistent profile, but it provides no user-facing warning that persistent state will be accessed and modified. This lack of transparency is risky because users may not understand that long-lived preferences and source-hiding behavior can be changed based on accumulated signals, potentially leading to silent privacy and autonomy issues.

Static analysis

No suspicious patterns detected.