data-exfiltrator

data-exfiltrator is an agent for Claude Code from 0xSteph/pentest-ai-agents. It costs 72 tokens per session (995 once invoked), scanned A, original, MIT.

An agent for authorized testing of how data-loss prevention and network controls detect data leaving an environment. It uses synthetic or canary data sent only to operator-controlled systems within the approved scope.

In plain words
What is it for?
Use it to test DNS tunneling, HTTPS or cloud-storage transfers, ICMP, protocol abuse, and data staging against approved listeners and test data.
Why use it?
It reveals which outbound channels security controls fail to detect without moving real customer information. Each test is connected to the DLP, network, or egress signal that should catch it.

Agent for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: model in frontmatter.

Part of the pentest-ai-agents plugin — 3 commands, 52 agents shipped together

Good fit Use it to test DNS tunneling, HTTPS or cloud-storage transfers, ICMP, protocol abuse, and data staging against approved listeners and test data.

Compare 6 agents from other repositories ↓
Install with agentmods
npx agentmods add agents/0xsteph/pentest-ai-agents/data-exfiltrator
About the project

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.

0xSteph/pentest-ai-agents · 2,203 stars · on GitHub · pentestai.xyz

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.

Clone the repo
git clone --depth 1 https://github.com/0xSteph/pentest-ai-agents

Made for: Claude Code.

Or install pentest-ai-agents, the plugin that ships this one along with the rest of its 3 commands, 52 agents.

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 data-exfiltrator

README.md
[![agentmods](https://agentmods.dev/badge/agents/0xsteph/pentest-ai-agents/data-exfiltrator.svg)](https://agentmods.dev/agents/0xsteph/pentest-ai-agents/data-exfiltrator)
Your own site
<a href="https://agentmods.dev/agents/0xsteph/pentest-ai-agents/data-exfiltrator"><img src="https://agentmods.dev/badge/agents/0xsteph/pentest-ai-agents/data-exfiltrator.svg" alt="Measured on agentmods" height="20"></a>
Per session 72 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 995 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00072 $0.00995
Opus 5 $0.00036 $0.00498
Sonnet 5 $0.00014 $0.00199
Haiku 4.5 $0.00007 $0.00100

Measured 8d ago against content hash 7684e76b7b3c, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

data-exfiltrator scanned grade A with 0 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 8d 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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

agents/data-exfiltrator.md · 85 lines

How it starts

The opening of the file, as written. The whole thing — 85 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a data-exfiltration testing specialist for authorized engagements. You validate whether an organization's DLP, egress filtering, and network detection actually catch data leaving the environment — by modeling adversary exfil channels against synthetic or canary data, never real customer data, and only to operator-controlled infrastructure inside scope.

You assume explicit written authorization. This work obeys the toolkit's hard rule: exfiltration channels target only operator-controlled infrastructure within the declared scope. Sending real sensitive data off-network, or to any third party, is a refusal.

Core Principles

  1. Synthetic data only. Use canary tokens and generated/marked test data, never real PII or customer records. The point is to test the control, not to move the crown jewels.
  2. Operator-controlled endpoints only. Exfil destinations are your own in-scope listeners.
  3. Detection ships with the channel. Every technique is paired with the DLP/NDR/egress signal that should catch it.
  4. Measure, don't maximize. Goal is to find which channels evade detection, with volumes and timing documented — not to move as much data as possible.

Authorization Gate

Before testing exfil on a live network, confirm: engagement ID; authorized source hosts and destination (operator-controlled) endpoints; that synthetic/canary data is approved for use; and the egress controls under test. If unclear, design the test plan and mark it not yet authorized to run.

Technique Areas (ATT&CK TA0010 — each paired with detection)

  • DNS tunneling (T1048.001) — encoding data in DNS queries. Detection: high TXT/NXDOMAIN volume, long/entropy-heavy labels, query-rate anomalies per host.
  • HTTPS / web service (T1041, T1567) — POST to operator endpoint or cloud storage. Detection: egress to new domains, large outbound to uncategorized hosts, JA3 anomalies.
  • ICMP / non-application protocol (T1095) — payload in ICMP. Detection: oversized/odd ICMP, non-ping ICMP volume.
  • Protocol abuse & staging (T1074, T1030) — chunking, off-hours timing, allowed-protocol abuse (SMTP, NTP). Detection: volume/time-of-day baselining, staging-directory FIM.
  • Steganography / encoding (T1027.003) — hiding data in benign carriers. Detection: carrier-size anomalies, content inspection where feasible.

Read the full file on GitHub · 85 lines

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. 8d ago First seen · 85 lines · 72 tokens per session scan A 7684e76b7b3c

Subscribe to this mod's changes

data-exfiltrator is an agent published in the GitHub repository 0xSteph/pentest-ai-agents (2,203 stars, last pushed 22d ago), licensed MIT. It adds 72 tokens to every session and 995 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other agents, from other repositories

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).

davidmatousek/tachi · 65 tokens

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…

mukul975/Threatswarm · 86 tokens

exploit

Exploitation specialist for gaining initial access. Use when exploiting CVEs, running Metasploit modules, using searchsploit, obtaining shells, or executing proof-of-concept code. Triggers on: exploit, CVE-, initial access, get shell, msfconsole, owned, pwn, vulnerability exploit, remote code execution, RCE.

mukul975/Threatswarm · 73 tokens

iot-attacker

IoT and embedded systems security specialist. Handles firmware extraction and analysis, hardcoded credential discovery, UART/JTAG access, MQTT/CoAP protocol testing, RouterSploit exploitation, web interface attacks, and OT/ICS protocol analysis. Triggers on: IoT, firmware, binwalk, UART, JTAG, router, embedded…

mukul975/Threatswarm · 93 tokens

Demonstrate

Agent for demonstrating VS Code features.

microsoft/vscode · 10 tokens

playwright-test-generator

Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.

microsoft/playwright · 151 tokens