ClawHub
Back to skill

Security audit

Decision Log

Security checks for vulnerabilities and agentic risk

Overview

This skill is a plain Markdown decision journal that creates local decision-log files, with no evidence of hidden code, remote execution, credential access, or destructive behavior.

Before installing, decide where decisions/ will live and avoid putting sensitive business, financial, client, or personal details in synced or public repositories. Review generated entries before saving, and narrow or skip the optional automatic triggers if you do not want routine spending or project mentions to prompt logging.

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 (4)

Vague Triggers

Medium
Confidence
94% confidence
Finding
The instruction triggers whenever the user mentions spending money on anything, which is a very broad condition that overlaps with ordinary conversation. The file does not provide constraints or negative examples to distinguish casual mentions from cases where the skill should actually activate.

Vague Triggers

Medium
Confidence
91% confidence
Finding
The phrase 'When I start a new project (create a project file or say "starting [X]")' is not specific enough about scope, and 'starting [X]' is common language that could occur outside intentional skill use. The document does not define boundaries or exclusions for what counts as a qualifying project start.

Vague Triggers

Medium
Confidence
90% confidence
Finding
The condition 'When I add a new tool or service subscription' is ambiguous because it does not define how the agent detects the event or distinguish substantive adoption from casual discussion, trials, or minor account changes. This can lead to unintended automatic entry creation without clear trigger boundaries.

Missing User Warnings

Low
Confidence
92% confidence
Finding
The skill explicitly instructs the agent to create and update files in `decisions/`, including an index and decision entries, but does not clearly warn the user that using the skill will modify the local filesystem and may store sensitive business, financial, partnership, or personal decision data in plain files. In this context the behavior is core to the skill rather than covert, so the risk is limited, but users could unintentionally persist sensitive information or allow broad automated file edits without informed consent.

Static analysis

No suspicious patterns detected.