ClawHub
Back to skill

Security audit

Kanban

Security checks for vulnerabilities and agentic risk

Overview

This Kanban skill is a local task-board helper that clearly discloses its file-based persistence and does not include network, credential, or executable behavior.

Install this only if you want a persistent local Kanban system. Expect it to create and update Markdown files under ~/kanban/ or a project .kanban directory, including board state, routing metadata, workspace paths, aliases, and logs. Prefer explicit-only activation or selected-project activation if you do not want planning conversations to automatically use Kanban state.

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 (3)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill explicitly instructs the agent to create board.md when missing and to initialize new project entries without requiring a clear user confirmation that local files will be created or modified. In an agent setting, silent writes to workspace or home-directory state can cause unintended persistence, overwrite user expectations, and broaden impact across sessions because the skill is designed to run automatically at conversation start.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The protocol directs the agent to perform automatic board repair, update registry metadata, and append to logs after writes, again without an explicit warning or approval flow for modifying existing files. Even if labeled 'non-destructive,' automatic repair can still alter user data, create misleading audit entries, and persist unintended state in shared or home-directory locations across projects and sessions.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The setup instructs the skill to activate on broad conversational cues such as 'always when task planning appears,' which can cause the agent to engage and persist state without clear user intent. In a skill that writes routing and memory files across projects and sessions, overly broad activation increases the risk of unintended file creation, cross-project contamination, and collection of user workflow data when the user did not explicitly ask for Kanban behavior.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.