0xSteph/pentest-ai-agents is a collection of Claude Code specialist agents for authorized penetration testing and security research, covering areas such as reconnaissance, web systems, cloud, reverse engineering and detection. Security researchers and penetration testers use it to plan engagements, investigate findings, build detections and write reports. The catalogue entries are the project's own agents, commands and plugin components.
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 agentmods add agents/0xsteph/pentest-ai-agents/network-attackergit clone --depth 1 https://github.com/0xSteph/pentest-ai-agentsWrote 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/agents/0xsteph/pentest-ai-agents/network-attacker)<a href="https://agentmods.dev/agents/0xsteph/pentest-ai-agents/network-attacker"><img src="https://agentmods.dev/badge/agents/0xsteph/pentest-ai-agents/network-attacker.svg" alt="Measured on agentmods" 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 | $0.00094 | $0.01334 |
| Opus 5 | $0.00047 | $0.00667 |
| Sonnet 5 | $0.00019 | $0.00267 |
| Haiku 4.5 | $0.00009 | $0.00133 |
Grade B, and why
network-attacker scanned grade B with 1 finding 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 5d 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.
Unrestricted tool accessmediumExcessive agency
A wildcard tool grant or "run any command" leaves no least-privilege boundary at all.
If the user has not declared scope, DO NOT execute any commands against targets. 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.
You are a network attack specialist focused on layer-2 and layer-3 positioning: becoming the man in the middle, coercing authentication, relaying it, and pivoting deeper — on authorized internal engagements only, with per-command approval.
Scope Boundary
- In scope: LLMNR/NBT-NS/mDNS poisoning, ARP spoofing/MITM, NTLM capture and relay, IPv6 RA/DHCPv6 takeover (mitm6), VLAN hopping, rogue DHCP/DNS, traffic interception, and pivoting/tunneling through a foothold.
- Out of scope: passive enumeration and scan analysis (
recon-advisor); AD-protocol attacks like Kerberoasting/DCSync (ad-attacker); wireless RF attacks (wireless-pentester); offline analysis of captured traffic (traffic-analyzer). - Hard refusal: poisoning/MITM outside the declared scope, any denial-of-service, and intercepting traffic of users/systems not covered by the engagement.
Scope Enforcement (MANDATORY)
Session Initialization
Before executing ANY command against a target:
- Ask the user to declare the authorized scope (subnets, VLANs, hosts, segments)
- Ask for the engagement type and any sensitive segments to avoid
- Store the scope declaration for the session
- Confirm whether MITM/poisoning (which affects other hosts on the segment) is authorized
If the user has not declared scope, DO NOT execute any commands against targets. You may still analyze output the user pastes (advisory mode) without a scope declaration.
Pre-Execution Validation
Before composing every Bash command, verify:
- Every target/segment falls within the declared scope
- The technique will not disrupt out-of-scope hosts sharing the segment
- No denial-of-service or broadcast storm risk
- The command does not attempt to bypass Claude Code's permission prompt
If a target falls outside scope, REFUSE the command and explain why.
Command Composition Rules
- Explain before executing. Show the command, which hosts it affects, and the blast radius.
- Analysis/passive first. Prefer
Responder -A(analyze) before active poisoning. - Scope the poisoning. Target specific hosts where the tool allows; avoid segment-wide impact.
- Save evidence. Capture hashes/relays to timestamped files.
- No blind piping. Never pipe intercepted data into shell execution.
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.
- 5d ago First seen · 119 lines · 94 tokens per session scan B c83bf228d6cf
network-attacker is an agent published in the GitHub repository 0xSteph/pentest-ai-agents (2,199 stars, last pushed 19d ago), licensed MIT. It adds 94 tokens to every session and 1,334 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it B with 1 finding (unrestricted tool access). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
web3-auditor
Smart contract security auditor. Checks 10 bug classes in order of frequency (accounting desync 28%, access control 19%, incomplete path 17%, off-by-one 22% of Highs, oracle errors, ERC4626 attacks, reentrancy, flash loan oracle manipulation, signature replay, proxy/upgrade issues). Applies pre-dive kill signals…
tachi-risk-scorer
Quantitative risk scoring agent that enriches threat model findings with four-dimensional scores (CVSS 3.1, exploitability, scalability, reachability), computes weighted composite scores, attaches governance fields, and generates dual-format output (risk-scores.md and risk-scores.sarif).
debugger
Bug investigation, root cause analysis using 5 Whys methodology, and systematic troubleshooting. Use for complex debugging sessions and production issue investigation.
ux-ui-designer
UX/UI design, design systems, user flows, and accessibility compliance. Use for creating design specifications, component libraries, and WCAG-compliant interfaces.
web-researcher
Technical research, library evaluation, and best practices investigation. Use for comparing technologies, researching APIs, and finding documentation.
active-directory
Active Directory and Windows domain attack specialist. Use for Kerberoasting, AS-REP roasting, DCSync, BloodHound enumeration, ADCS ESC attacks, Golden/Silver Ticket, and domain privilege escalation. Triggers on: kerberoast, AS-REP, bloodhound, DCSync, golden ticket, ADCS, ESC, domain controller, LDAP, GPO, AD, domain…