ClawHub
Back to skill

Security audit

WorkBuddy 自定义模型能力位排查（读图/工具/推理）

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed WorkBuddy troubleshooting tool that probes configured model APIs and can update local model capability flags when explicitly requested.

Install only if you want Codex to help test and repair WorkBuddy custom model capability flags. Run the probe read-only first, use trusted model endpoints, avoid passing private images with --image unless you trust that endpoint, and review the reported changes before using --apply or --add.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (6)

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
97% confidence
Finding

The skill contains explicit commands and workflows that read and write local files, invoke shell/Python, inspect environment-dependent paths, and make outbound network requests to model endpoints, but it declares no tool scope or permissions. That mismatch is dangerous because an agent may execute these actions without clear least-privilege boundaries or user awareness, increasing the chance of unintended file modification, credential exposure, or network exfiltration.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest name/display text and long-form description are entirely in Chinese and present the skill as intended for use in that language, but there is no statement that users may interact in other languages or choose their preferred locale. Under the policy, a language constraint without opt-in or justification is a natural-language policy violation.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 64)May include surrounding context.

md
# 加一个全新模型：先追加条目，再探测，再按探测结果填能力位
python probe_model.py --add --model my-new-model \
  --url https://api.example.com/chat/completions --key sk-xxx \
  --name "My Model" --vendor "Example"

# 常用组合

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The embedded vision prompt instructs the model to answer in a fixed English format, and the tool prompt is also written only in English. This is a locale/language constraint expressed in natural language with no user choice or opt-in, which matches the language policy violation category.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · test_vision_stability.py (reported line 17)May include surrounding context.

python
import probe_model as pm

MODEL = os.environ.get("PROBE_MODEL", "deepseek-v4-flash")
URL = os.environ.get("PROBE_URL", "https://api.deepseek.com/chat/completions")
# ⚠ key 只能从环境变量读。此前这个文件里硬编码过一枚真实 key，
# skill 一旦发布就等于公开泄露 —— 已改为必须显式提供。
KEY = os.environ.get("DEEPSEEK_API_KEY") or os.environ.get("PROBE_KEY")

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
81% confidence
Finding

This code overwrites the user's configuration file and persists backup files on disk. Although the CLI help mentions writing back when using --apply/--add, the code path performing the write has no confirmation prompt and no explicit user-facing warning immediately before the file modification, which reduces disclosure for a safety-relevant file write.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.