100Critical
This skill is an exploitation toolkit enabling arbitrary
Behavior Analysis
Description verified — behavior matches claim
Claims to do
CTF Binary Exploitation (Pwn): Quick reference for binary exploitation (pwn) CTF challenges. Each technique has a one-liner here; see supporting files for full details.
Actually does
The skill provides installation commands for binary exploitation tools (`pip`, `apt`, `brew`, `gem`), and offers quick-start commands for binary analysis (`checksec`, `file`, `readelf`), gadget finding (`ROPgadget`, `ropper`, `one_gadget`), debugging (`gdb`), and libc identification (`./libc-database/find`). It also includes a `bash` command for race condition exploitation via `nc` and references numerous local markdown files for detailed exploitation techniques. It is configured to use `Bash`, `Read`, `Write`, `Edit`, `Glob`, `Grep`, `Task`, `WebFetch`, and `WebSearch` tools.
Skill Info
Nameljagiello/ctf-skills/ctf-pwn
Registrygithub
Version1af14f9030fe
PURLpkg:github/ljagiello/ctf-skills@1af14f9030fe?skill=ctf-pwn
Stars1,876
Installs3,000
Source
SHA-2567a903ffd7013...
SHA-188dffd56b2b6...
MD533ef3348d258...
TLSH5c821a637621...
ssdeep384:1J5kQxBn...
Size19,069 bytes
Install
This skill has been flagged as potentially malicious. Review the findings below before installing.
Skills.shDocs
npx skills add https://github.com/ljagiello/ctf-skillsAssessments (2)
AI Agent Traps ↗Shell command execution function detectedcriticalcommand execution
Shell command execution function detected
100% confidenceline 53
system(
Shell command execution function detectedcriticalcommand execution
Shell command execution function detected
100% confidenceline 160
system(
Shell command execution function detectedcriticalcommand execution
Shell command execution function detected
100% confidenceline 166
system(
Shell command execution function detectedcriticalcommand execution
Shell command execution function detected
100% confidenceline 194
system(
Broad Command Execution and System Toolinghighcommand execution
The skill explicitly allows broad command execution via `Bash` and `Python`, and instructs the installation and use of numerous powerful system and exploitation tools, enabling arbitrary code execution.
100% confidenceline 6
allowed-tools: Bash Read Write Edit Glob Grep Task WebFetch WebSearch, pip install pwntools..., apt install gdb..., python3 -c "from pwn import *; print(cyclic(200))"
Privilege Escalation and Sandbox Escape Capabilitieshighprivilege escalation
The skill's description and detailed content explicitly cover techniques for privilege escalation, kernel exploitation, seccomp bypass, and sandbox escape, enabling an agent to gain elevated access or break out of confined environments.
100% confidenceline 194
description: ...privilege escalation, ...kernel exploitation, seccomp bypass, sandbox escape..., kernel.md, sandbox-escape.md, CPU emulator eval injection
Data Exfiltration Techniques Describedmediumdata exfiltration
The skill describes techniques such as 'time-based blind shellcode exfiltration' and arbitrary read primitives, providing methods for an agent to extract data from a target system.
90% confidence
time-based blind shellcode exfiltration (rop-advanced.md), arbitrary read/write GOT overwrite to shell (advanced-exploits-3.md)
Network Exploitation via Netcatmediumcommand execution
The skill includes direct examples of using `netcat` (`nc`) for network interaction, which is a common method for establishing reverse shells or exfiltrating data in exploitation scenarios.
100% confidenceline 104
bash -c '{ echo "cmd1"; echo "cmd2"; sleep 1; } | nc host port'Oversight Evasion Framing as 'CTF Challenges'lowoversight evasion
The skill's content is consistently framed as being for 'CTF challenges,' which could serve as an oversight evasion tactic to normalize or justify the use of inherently malicious exploitation techniques.
80% confidenceline 3
name: ctf-pwn, description: Provides binary exploitation techniques for CTF challenges.
Badge
Markdown
[](https://mondoo.com/ai-agent-security/skills/github/ljagiello/ctf-skills/ctf-pwn)HTML
<a href="https://mondoo.com/ai-agent-security/skills/github/ljagiello/ctf-skills/ctf-pwn"><img src="https://mondoo.com/ai-agent-security/api/badge/github/ljagiello/ctf-skills/ctf-pwn.svg" alt="Mondoo Skill Check" /></a>Image URL
https://mondoo.com/ai-agent-security/api/badge/github/ljagiello/ctf-skills/ctf-pwn.svgSecure your AI agents
Skills can read files, run commands, and access credentials. Mondoo helps organizations manage the security risks of AI agent skills across their entire fleet.
- Continuous scanning of skills across all registries
- Policy enforcement before skills reach your agents
- Integration with your existing security workflow