Dockerfile

# ──────────────────────────────────────────────────────────────────────────────
# HexStrike AI on Hugging Face Spaces — Debian bookworm base (no Kali)
#
# Installs the full HexStrike-recommended toolset in cache-friendly layers.
# Final image ~6–9 GB, well under HF Spaces' 50 GB ephemeral root quota.
#
# ⚠️ Non-root limitations on HF Spaces (UID 1000, no CAP_NET_RAW):
#   * masscan, responder, rustscan-SYN, nmap -sS/-sU/-O — installed but will
#     fail at runtime with "permission denied". They're included so HexStrike
#     doesn't error on "tool not found"; just don't have the LLM call them.
#     Use nmap -sT (TCP connect) instead.
# ──────────────────────────────────────────────────────────────────────────────
FROM python:3.11-slim-bookworm

ENV DEBIAN_FRONTEND=noninteractive \
    PYTHONUNBUFFERED=1 \
    PIP_NO_CACHE_DIR=1 \
    PIP_BREAK_SYSTEM_PACKAGES=1 \
    LANG=C.UTF-8 \
    HF_HOME=/home/user/.cache/huggingface \
    GOPATH=/opt/go \
    PATH="/opt/go/bin:/usr/local/bin:/root/.local/bin:${PATH}"

# ── Enable contrib + non-free (nikto and friends live there) ────────────────
RUN printf '%s\n' \
      'deb http://deb.debian.org/debian bookworm main contrib non-free non-free-firmware' \
      'deb http://deb.debian.org/debian bookworm-updates main contrib non-free non-free-firmware' \
      'deb http://deb.debian.org/debian-security bookworm-security main contrib non-free non-free-firmware' \
      > /etc/apt/sources.list

# ── Layer 1: build essentials + general utilities ───────────────────────────
RUN apt-get update && apt-get install -y --no-install-recommends \
        ca-certificates curl wget git jq unzip xz-utils tar gzip gnupg \
        build-essential pkg-config \
        python3-dev python3-pip pipx \
        ruby ruby-dev rubygems \
        default-jre-headless \
    && rm -rf /var/lib/apt/lists/*

# ── Layer 2: network & reconnaissance (apt) ─────────────────────────────────
# Dropped from this list because they don't exist in Debian bookworm main:
#   * amass     → installed below from upstream Go release
#   * responder → installed below via pipx (Responder is on PyPI)
RUN apt-get update && apt-get install -y --no-install-recommends \
        nmap masscan \
        dnsutils whois \
        dnsenum fierce \
        netcat-openbsd \
        smbclient \
    && rm -rf /var/lib/apt/lists/*

# ── Layer 2b: amass from upstream Go release ────────────────────────────────
ARG AMASS_VERSION=4.2.0
RUN curl -fsSL -o /tmp/amass.zip \
      "https://github.com/owasp-amass/amass/releases/download/v${AMASS_VERSION}/amass_Linux_amd64.zip" \
 && unzip -o /tmp/amass.zip -d /tmp/amass \
 && mv /tmp/amass/amass_Linux_amd64/amass /usr/local/bin/amass \
 && chmod +x /usr/local/bin/amass \
 && rm -rf /tmp/amass /tmp/amass.zip

# ── Layer 3: web app security (apt) ─────────────────────────────────────────
RUN apt-get update && apt-get install -y --no-install-recommends \
        gobuster ffuf dirb \
        nikto whatweb wafw00f \
        sqlmap \
    && rm -rf /var/lib/apt/lists/*

# ── Layer 4: password & auth tools (apt) ────────────────────────────────────
RUN apt-get update && apt-get install -y --no-install-recommends \
        hydra medusa patator \
        john hashcat \
    && rm -rf /var/lib/apt/lists/*

# ── Layer 5: binary analysis / RE (apt) ─────────────────────────────────────
RUN apt-get update && apt-get install -y --no-install-recommends \
        gdb \
        binwalk foremost steghide \
        binutils file \
        libimage-exiftool-perl \
        checksec \
    && rm -rf /var/lib/apt/lists/*

# ── Layer 6: Chromium for HexStrike's BrowserAgent ──────────────────────────
RUN apt-get update && apt-get install -y --no-install-recommends \
        chromium chromium-driver \
        fonts-liberation libnss3 \
    && rm -rf /var/lib/apt/lists/*
ENV CHROME_BIN=/usr/bin/chromium \
    CHROMEDRIVER=/usr/bin/chromedriver

# ── Layer 7: ProjectDiscovery + Rust binaries (GitHub releases) ─────────────
ARG NUCLEI_VERSION=3.3.7
ARG SUBFINDER_VERSION=2.6.6
ARG HTTPX_VERSION=1.6.9
ARG KATANA_VERSION=1.1.0
ARG NAABU_VERSION=2.3.2
ARG FEROXBUSTER_VERSION=2.11.0
ARG RUSTSCAN_VERSION=2.3.0

RUN set -eux; \
    cd /tmp; \
    for spec in \
        "nuclei|${NUCLEI_VERSION}" \
        "subfinder|${SUBFINDER_VERSION}" \
        "httpx|${HTTPX_VERSION}" \
        "katana|${KATANA_VERSION}" \
        "naabu|${NAABU_VERSION}" \
    ; do \
        name=${spec%|*}; ver=${spec#*|}; \
        curl -fsSL -o pkg.zip \
          "https://github.com/projectdiscovery/${name}/releases/download/v${ver}/${name}_${ver}_linux_amd64.zip"; \
        unzip -o pkg.zip -d /usr/local/bin/ "${name}"; \
        chmod +x "/usr/local/bin/${name}"; \
        rm pkg.zip; \
    done; \
    curl -fsSL -o ferox.zip \
      "https://github.com/epi052/feroxbuster/releases/download/v${FEROXBUSTER_VERSION}/x86_64-linux-feroxbuster.zip"; \
    unzip -o ferox.zip -d /usr/local/bin/ feroxbuster && chmod +x /usr/local/bin/feroxbuster && rm ferox.zip; \
    curl -fsSL -o /tmp/rustscan.deb \
      "https://github.com/bee-san/RustScan/releases/download/${RUSTSCAN_VERSION}/rustscan_${RUSTSCAN_VERSION}_amd64.deb"; \
    (dpkg -i /tmp/rustscan.deb || (apt-get update && apt-get install -fy && rm -rf /var/lib/apt/lists/*)); \
    rm /tmp/rustscan.deb

# ── Layer 8: Python-based security tools (pipx, isolated venvs) ─────────────
# pipx gives each tool its own venv so conflicting deps can't fight.
# `|| echo warn` keeps a transient PyPI hiccup from killing the whole build.
ENV PIPX_HOME=/opt/pipx PIPX_BIN_DIR=/usr/local/bin
RUN pipx ensurepath && \
    for tool in \
        dirsearch \
        arjun \
        paramspider \
        theharvester \
        autorecon \
        enum4linux-ng \
        netexec \
        impacket \
        volatility3 \
        Responder \
        hashid \
    ; do \
        pipx install --pip-args="--no-cache-dir" "$tool" || echo "[warn] pipx failed for $tool"; \
    done

# ── Layer 9: dalfox (Go binary via official installer) ──────────────────────
RUN curl -fsSL https://raw.githubusercontent.com/hahwul/dalfox/main/scripts/install.sh \
      | bash -s -- -b /usr/local/bin || echo "[warn] dalfox install failed"

# ── Layer 10: Ruby gems: evil-winrm, wpscan ─────────────────────────────────
RUN gem install --no-document evil-winrm wpscan || echo "[warn] gem install partial failure"

# ── Layer 11: cloud / container security ────────────────────────────────────
RUN curl -fsSL https://aquasecurity.github.io/trivy-repo/deb/public.key \
      | gpg --dearmor -o /usr/share/keyrings/trivy.gpg && \
    echo "deb [signed-by=/usr/share/keyrings/trivy.gpg] https://aquasecurity.github.io/trivy-repo/deb bookworm main" \
      > /etc/apt/sources.list.d/trivy.list && \
    apt-get update && apt-get install -y --no-install-recommends trivy && \
    rm -rf /var/lib/apt/lists/*

RUN pipx install prowler    || echo "[warn] prowler install failed"
RUN pipx install scoutsuite || echo "[warn] scoutsuite install failed"
RUN pipx install kube-hunter || echo "[warn] kube-hunter install failed"

ARG KUBEBENCH_VERSION=0.10.1
RUN curl -fsSL -o /tmp/kb.tgz \
      "https://github.com/aquasecurity/kube-bench/releases/download/v${KUBEBENCH_VERSION}/kube-bench_${KUBEBENCH_VERSION}_linux_amd64.tar.gz" \
    && tar -xzf /tmp/kb.tgz -C /usr/local/bin/ kube-bench \
    && chmod +x /usr/local/bin/kube-bench && rm /tmp/kb.tgz

RUN git clone --depth 1 https://github.com/docker/docker-bench-security.git \
      /opt/docker-bench-security || echo "[warn] docker-bench clone failed"

# ── Layer 12: SecLists wordlists + HexStrike-expected paths ─────────────────
RUN mkdir -p /usr/share/wordlists/dirb \
 && (curl -fsSL -o /usr/share/wordlists/dirb/common.txt \
        https://raw.githubusercontent.com/danielmiessler/SecLists/master/Discovery/Web-Content/common.txt \
  || curl -fsSL -o /usr/share/wordlists/dirb/common.txt \
        https://raw.githubusercontent.com/3ndG4me/KaliLists/master/dirb/common.txt) \
 && test -s /usr/share/wordlists/dirb/common.txt

# ── Layer 13: non-root user (HF Spaces requires UID 1000) ───────────────────
RUN useradd -m -u 1000 user \
 && mkdir -p /home/user/.local/bin \
 && chown -R user:user /home/user
USER user
WORKDIR /home/user/app

# Root-installed pipx binaries are in /usr/local/bin (PIPX_BIN_DIR) already.
# Add user's local pip bin too.
ENV PATH="/home/user/.local/bin:/usr/local/bin:${PATH}"

# ── Layer 14: HexStrike source + its own Python deps + mcp-proxy ────────────
RUN git clone --depth 1 https://github.com/0x4m4/hexstrike-ai.git hexstrike

COPY --chown=user:user requirements.txt .

RUN pip install --user --no-cache-dir flask gunicorn psutil requests aiohttp bs4 \
 && pip install --user --no-cache-dir -r requirements.txt \
 && pip install --user --no-cache-dir -r hexstrike/requirements.txt || true

# ── Layer 15: entrypoint ────────────────────────────────────────────────────
COPY --chown=user:user entrypoint.sh app.py ./
RUN chmod +x entrypoint.sh
RUN chmod +x app.py

EXPOSE 7860
CMD ["./entrypoint.sh"]