Security audit

Context Memoize Cache

Security checks for vulnerabilities and agentic risk

Overview

This is a small local context-cache skill with disclosed persistence and a scoped cache deletion command that deserves caution but does not show malicious behavior.

Before installing, understand that saved fragments may be loaded into future sessions. Do not cache secrets, credentials, or untrusted instructions unless you intend them to persist, and only run the clear-cache command when you intentionally want to delete the saved cache directory.

SkillSpector

Vulnerability Patterns

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
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The skill documents a destructive cache-clearing command using `rm -rf ~/.openclaw/context-cache/` without any warning, confirmation step, or safer alternative. Even though the path is scoped to a cache directory, normalizing destructive deletion in an agent skill increases the chance of accidental data loss if invoked automatically, copied incorrectly, or modified unsafely.

Static analysis

Detected: suspicious.destructive_delete_command

Documentation contains a destructive delete command without an explicit confirmation gate.

Warn

Code: suspicious.destructive_delete_command
Location: SKILL.md:19