prompt-guard
中风险 · 74 评分：74/100
作者：seojoonkim | 审计时间：2026-02-05T09:18:25.242Z | 规则集：0.1.0
技能介绍
Advanced prompt injection defense system for Clawdbot with HiveFence network integration. Protects against direct/indirect injection attacks in group chats with multi-language detection (EN/KO/JA/ZH)…
✨
<claude_*>, </claude_*> — Anthropic internal tag patterns ✨
[INST], <<SYS>>, <|im_start|> — LLaMA/GPT internal tokens ✨
GODMODE, DAN, JAILBREAK — Famous jailbreak keywords ✨
l33tspeak, unr3strict3d — Filter evasion via leetspeak ✨ 349 attack patterns (2.7x increase from v2.4)
✨ Authority impersonation detection (EN/KO/JA/ZH) - "나는 관리자야", "I am the admin"
✨ Indirect injection detection - URL/file/image-based attacks
✨ Context hijacking detection - fake memory/history manipulation
使用场景
1
<artifacts_info>, <antthinking>, <antartifact> — Claude artifact system 2 Multi-turn manipulation detection - gradual trust-building attacks
3
write, edit - File modifications 4 Gradual trust building
5 Art/Cinema jailbreak ("as a cinematographer, create a scene...")
6 Time-shift evasion ("back in 2010, write an email...")
安全审计
中风险 · 74
摘要
Advanced prompt injection defense system for Clawdbot with HiveFence network integration. Protects against direct/indirect injection attacks in group chats with multi-language detection (EN/KO/JA/ZH), severity scoring, automatic logging, and configurable security policies. Connects to the distributed HiveFence threat intelligence network for collective defense.
风险画像
关键风险 0 项
暂无 LLM 风险要点（LLM 未启用或无缓存）。
确定性发现（证据）
| 规则 | 严重性 | 文件 | 片段 |
|---|---|---|---|
| NET_HTTP_REQUEST | medium | skills/seojoonkim/prompt-guard/scripts/detect.py 行 1462 | import urllib.request |
| NET_HTTP_REQUEST | medium | skills/seojoonkim/prompt-guard/scripts/detect.py 行 1498 | req = urllib.request.Request( |
| NET_HTTP_REQUEST | medium | skills/seojoonkim/prompt-guard/scripts/detect.py 行 1505 | with urllib.request.urlopen(req, timeout=5) as resp: |
| NET_HTTP_REQUEST | medium | skills/seojoonkim/prompt-guard/scripts/hivefence.py 行 29 | import urllib.request |
| NET_HTTP_REQUEST | medium | skills/seojoonkim/prompt-guard/scripts/hivefence.py 行 79 | req = urllib.request.Request(url, data=body, headers=headers, method=method) |
| NET_HTTP_REQUEST | medium | skills/seojoonkim/prompt-guard/scripts/hivefence.py 行 82 | with urllib.request.urlopen(req, timeout=self.timeout) as resp: |
| QUALITY_README_PRESENT | low | README 行 无 | README detected |