Deploy hardened containers with declarative isolation

Nucleus

Extremely lightweight, security-hardened, declarative container runtime for agents and production services

Nucleus is a minimalist container runtime for Linux. It provides isolated execution environments using Linux kernel primitives without the overhead of traditional container runtimes. For production services, it is designed around a fully declarative model: Nix builds the root filesystem or image, the NixOS module declares the service, and Nucleus mounts pinned, reproducible runtime inputs.

Nucleus supports three operating modes:

• Agent mode (default) - ephemeral, fast-startup sandboxes for AI agent workloads
• Strict agent mode - fail-closed isolation for ephemeral agent workloads without requiring production rootfs, health checks, sd_notify, or NixOS service semantics
• Production mode - strict isolation for long-running, network-bound NixOS services with declarative configuration, reproducible Nix-built root filesystems/images, egress policy enforcement, health checks, and systemd integration

Production deployments are built to be:

• Fully declarative - service topology, runtime settings, mounted rootfs, and optional images are defined up front instead of assembled imperatively at deploy time
• Nix-native - first-class NixOS module support plus nucleus.lib.mkRootfs and nucleus.lib.mkImage for minimal service closures
• Reproducible - flake-based builds, pinned store paths, rootfs attestation, and image signatures keep runtime inputs stable and auditable

Benchmarks

Cold Start

Runtime	Startup Time
Nucleus	12 ms
Docker	~500 ms

PostgreSQL 18 (pgbench, 8 clients, 30s, scale 50, 3 runs)

The native runtime stays within run-to-run noise of bare metal, while the gVisor
runtime pays a consistent ~62% tax from the Sentry's per-syscall user-space
emulation. Occasional wins over bare metal should be treated as benchmark
noise rather than a guaranteed speedup.

SELECT-only (read-heavy)

Environment	I/O Method	Avg TPS	Avg Latency	vs Baremetal
Baremetal	worker	170,411	0.047 ms	baseline
Baremetal	io_uring	174,026	0.046 ms	+2.1%
Nucleus	worker	169,577	0.047 ms	−0.5%
Nucleus	io_uring	172,482	0.046 ms	+1.2%
gVisor	worker	65,683	0.122 ms	−61.5%

TPC-B (mixed read/write)

Environment	I/O Method	Avg TPS	Avg Latency	vs Baremetal
Baremetal	worker	4,341	1.845 ms	baseline
Baremetal	io_uring	4,516	1.772 ms	+4.0%
Nucleus	worker	4,583	1.746 ms	+5.6%
Nucleus	io_uring	4,607	1.736 ms	+6.1%
gVisor	worker	1,655	4.834 ms	−61.9%

Measured on Linux 6.18 x86_64 with PostgreSQL 18.4. All three environments run
fully unprivileged (Nucleus rootless with --userns keep-id, no sudo);
the harness bind-mounts a host pgdata and uses host networking, so it
measures the steady-state cost of each isolation layer rather than image
unpacking or cold start. gVisor's Sentry does not implement io_uring, so it
is measured with io_method=worker only. Full results:
benches/pg18_io/results/

Run it yourself (rootless):

nix shell nixpkgs#postgresql_18 -c \
  env ROOTLESS=1 SCALE=50 CLIENTS=8 DURATION=30 RUNS=3 \
  bash benches/pg18_io/bench.sh

Add GVISOR_PLATFORM=systrap if /dev/kvm is unavailable, or SKIP_GVISOR=1
to measure baremetal + native only.

Why Nucleus?

• Declarative by default for services - Production deployments are defined in NixOS and TOML rather than stitched together with ad hoc runtime scripting
• Deep Nix integration - First-class NixOS module, mkRootfs, mkImage, and Nix store closures for minimal, locked-down service roots
• Reproducible service builds - Flake-based packaging, pinned inputs, rootfs attestation, and image signatures make runtime state auditable and repeatable
• Zero-overhead isolation - Direct use of cgroups, namespaces, pivot_root, capabilities, seccomp, and Landlock
• Memory-backed filesystems - Container disk mapped to tmpfs, pre-populated with agent context
• gVisor integration - Optional application kernel for enhanced security, including networked service mode
• OCI runtime-spec subset for gVisor - Generates OCI bundle/config data for runsc, including process identity, mounts, namespaces, seccomp, hooks, and cgroup path wiring
• Detached mode - Run containers in the background as systemd transient services with --detach, managed via nucleus stop/logs/attach
• Production service support - Declarative NixOS module, egress policies, credential-broker egress, health checks, secrets mounting, sd_notify, and journald integration
• Explicit workload identity - Native and gVisor runtimes can drop to a configured uid/gid plus supplementary groups after privileged setup
• Minimal rootfs - Replace host bind mounts with a purpose-built Nix store closure or Nix-built image for production services
• Local image snapshots - Commit native overlay-backed containers to signed, thin image directories, then verify, inspect, and run them later
• External security policies - Per-service seccomp profiles (JSON), capability policies (TOML), and Landlock rules (TOML) with SHA-256 pinning
• Seccomp profile generation - Trace mode records syscalls, then nucleus seccomp generate creates a minimal allowlist profile
• Multi-container topologies - Compose-equivalent TOML format with dependency DAG, reconciliation, and NixOS systemd integration
• Integrity & audit controls - Structured audit log, machine-readable lifecycle event streams, context hashing, rootfs attestation, image signatures, seccomp deny logging, mount flag verification, and kernel lockdown assertions
• Structured telemetry - Optional OpenTelemetry export for container lifecycle tracing
• Linux-native - Runs on standard Linux and NixOS

Relationship to Docker

Nucleus is not a drop-in Docker replacement, nor a strict subset of Docker.
The feature sets overlap, but each tool does things the other does not. Nucleus is
a hardened sandbox runtime (closer in spirit to runc/gVisor) that also does
lightweight, declarative single-host orchestration. It drops Docker's build DSL,
registry, and distribution workflow in exchange for deeper isolation, policy,
and reproducibility. Local signed image snapshots are available, but they are
not Docker/OCI images.

Capability	Docker	Nucleus
Root filesystem	Layered image (union mount)	tmpfs directory (agent), Nix closure (production), or overlay-backed Nix closure for snapshots
Images / Dockerfile / registry	Yes	Signed local thin snapshots and Nix-built image manifests; no Dockerfile, registry, pull/push, or OCI image spec
Persistent storage	Named volumes + storage drivers	Ephemeral tmpfs; persistence only via explicit --volume binds
Architecture	dockerd daemon + socket API	Single binary, direct fork/exec; detached = systemd transient unit
Networking	CNI plugins, overlay networks	none / host / bridge only
Orchestration	Compose, Swarm	nucleus compose (single-host TOML DAG over systemd)
Default egress	Allow-all outbound	Deny-by-default; allow per CIDR/domain via namespace iptables
Filesystem ACLs	AppArmor/SELinux profiles	Landlock LSM, per-service, irreversible
gVisor	Optional add-on runtime	First-class integrated runtime with explicit network modes
Security policies	Bundled defaults	Externalized seccomp/caps/Landlock, SHA-256 pinned + trace-generated
Reproducibility	Image digests	Nix closures, rootfs attestation, image signatures, first-class NixOS module
Verification	-	TLA+ specs + model-based tests across subsystems
Default hardening	~300 syscalls, some caps kept	All caps dropped, small seccomp allowlist, up to 8 namespaces

If your mental model is "run my Docker image instead of docker run," it will
not fit: there is no Dockerfile, registry, pull/push lifecycle, or implicit
persistent state. Nucleus images are local signed snapshots or Nix-built
manifests over Nix rootfs closures. If your model is "run untrusted or ephemeral
workloads with stronger, auditable isolation," that is the target.

Architecture

Nucleus leverages Linux kernel isolation primitives:

• Namespaces - PID, mount, network, UTS, IPC, user, cgroup, and optional time isolation
• cgroups v2 - Resource limits (CPU, memory, PIDs, I/O)
• pivot_root - Filesystem isolation (chroot fallback available in agent mode only)
• Capabilities - All capabilities dropped by default, or configured via TOML policy file (irreversible)
• seccomp - Syscall whitelist filtering with per-service JSON profiles and trace-based generation (irreversible)
• Landlock - Path-based filesystem access control via hardcoded defaults or TOML policy file (Linux 5.13+)
• gVisor - Optional application kernel (runsc) with none, bridge handoff, and explicit gvisor-host network modes
• OCI bundle generation - Emits OCI config.json plus bundle layout for gVisor, including process.user, lifecycle hooks, seccomp, resource limits, and namespace mappings
• Image snapshots - Local signed manifests with optional overlay diffs rooted in attested Nix rootfs closures
• PID 1 init - Mini-init supervisor in production mode for zombie reaping and signal forwarding
• In-memory secrets - Dedicated tmpfs at /run/secrets with volatile zeroing of source buffers
• Mount audit - Post-setup verification of mount flags in production mode

Container filesystem is backed by tmpfs and either populated with context files (agent mode) or mounted from a pre-built Nix rootfs closure (production mode). Snapshot workflows can mount that Nix rootfs with a writable native overlay and commit the overlay upperdir as a signed local image. That lets services run from declaratively built, reproducible filesystem inputs instead of inheriting mutable host state.

Platform Support

• Linux (kernel 6.x+) on x86_64
• NixOS (first-class NixOS module support)
• Not supported: macOS, Windows, BSDs, 32-bit Linux

Installation

cargo install nucleus-container

Or via Nix (recommended for reproducible builds and NixOS integration):

nix run github:wiggum-cc/nucleus

The Cargo package name is nucleus-container; it installs the nucleus binary. The repository itself is packaged as a Nix flake, so nix run, nix build, and the NixOS module all share the same pinned inputs.

Recent Features

• GPU passthrough - --gpu auto|nvidia|amd|intel|all binds host GPU device nodes into /dev, installs a cgroup v2 device allowlist (BPF_PROG_TYPE_CGROUP_DEVICE), relaxes the seccomp ioctl filter for vendor driver ioctls, and bind-mounts driver support files. Explicit device selection, NVIDIA env vars, and a programmatic gpu config document field are supported. See spec/gpu-passthrough.md.
• Local signed image snapshots - Native overlay-backed containers can be committed, verified, inspected, loaded, and run as thin image directories over a Nix rootfs base.
• Privilege drop for services - --user, --group, and --additional-group now apply a real post-setup workload identity in both the native runtime and gVisor.
• Ownership-aware secrets and writable paths - Production secret staging and NixOS createHostPath = true defaults now align file ownership with the configured workload user/group.
• OCI bundle identity support - Generated gVisor OCI configs now carry process.user including supplementary groups, alongside namespaces, mounts, resource limits, seccomp, hooks, and cgroupsPath.
• Probe execution under workload identity - Exec-based health and readiness probes now run as the configured service account instead of implicitly as root.
• Systemd/NixOS service integration improvements - The module exposes user, group, and supplementaryGroups, and packaged Nix usage includes gvisor in the flake/dev shell path.

Usage

Agent Mode (default)

# Run agent in isolated container with pre-populated context
nucleus run --context ./agent-context/ -- /usr/bin/agent

# Specify resource limits
nucleus run --memory 512M --cpus 2 --context ./ctx/ -- ./agent

# Name your container
nucleus run --name my-agent --context ./ctx/ -- ./agent

# Use gVisor for enhanced isolation
nucleus run --runtime gvisor --context ./ctx/ -- ./agent

# Rootless mode
nucleus run --rootless -- /bin/sh

# Rootless system services (e.g. PostgreSQL) that refuse euid 0:
# --userns keep-id maps your uid to itself so host-owned bind mounts just work
# (requires /etc/subuid + /etc/subgid, exactly like Docker/Podman rootless).
nucleus create --userns keep-id --user "$(id -u)" --group "$(id -g)" \
  -v ./pgdata:/pgdata -- postgres -D /pgdata
# --user <non-zero> alone auto-selects keep-id when /etc/subuid is configured.
# --userns auto     Podman/Docker default (workload uid in the subuid range)
# --userns nomap    historic (only container root usable)
# --uidmap/--gidmap explicit container:host:size mappings (Podman syntax)

# Optional networking
nucleus run --network host --allow-host-network -- curl https://example.com
nucleus run --network bridge -p 8080:80 -- ./server
nucleus run --network bridge -p 127.0.0.1:8080:80 -- ./server
nucleus run --rootless --network bridge -- ./client
nucleus run --network bridge --nat-backend userspace -- ./client

# Context streaming (bind mount for instant access)
nucleus run --context ./large-dir/ --context-mode bind -- ./agent

# Integrity and audit hardening
nucleus run --context ./ctx/ --verify-context-integrity --seccomp-log-denied -- ./agent

# Environment variables
nucleus run -e DEBUG=1 -- ./agent

# Sensitive environment variables without argv exposure
printf '{"OPENAI_API_KEY":"..."}' | nucleus run --env-fd 3 3<&0 -- ./agent

# Pass sensitive values via --secret (mounted in-memory at /run/secrets)
nucleus run --secret /path/to/api-key:/run/secrets/api_key -- ./agent

# Run a coding agent against a stable /workspace cwd
nucleus run \
  --workspace "$PWD" \
  --workspace-mode bind-rw \
  --workspace-exec \
  -- ./agent

# Mount provider CLI config under the private home directory
nucleus run \
  --provider-config-ro "$HOME/.aws:.aws" \
  --provider-config-rw "$HOME/.config/gh:.config/gh" \
  -- ./agent

# Run an agent with a pinned provider/toolchain rootfs instead of host runtime binds
nucleus run \
  --service-mode mitos-agent \
  --agent-toolchain-rootfs /nix/store/...-nucleus-agent-toolchain-rootfs \
  --workspace "$PWD" \
  --workspace-exec \
  -- codex

Programmatic Launch Config

nucleus run accepts the same command as nucleus create. Programmatic callers
that need a stable launch contract can provide the whole request as JSON or TOML
instead of constructing a long argv list:

nucleus run --config ./agent.nucleus.toml
nucleus run --config ./agent.nucleus.json
nucleus run --config-fd 3 3<./agent.nucleus.json

Config mode owns the launch request: put the workload command and all sandbox
options in the config document rather than mixing them with per-option CLI flags.
The schema uses the long CLI option names converted to snake_case:

name = "mitos-agent"
workspace = "/home/dev/project"
workspace_mode = "bind-rw"
workspace_exec = true
workdir = "/workspace"
runtime = "gvisor"
service_mode = "strict-agent"
agent_toolchain_rootfs = "/nix/store/...-nucleus-agent-toolchain-rootfs"
memory = "1G"
cpus = 2.0
pids = 512
command = ["./agent", "--stdio"]

env_vars = ["RUST_LOG=info"]
seccomp_log_denied = true

Workspace

--workspace <host-path> mounts the host project tree at /workspace. The
process cwd defaults to /workspace via --workdir /workspace.

--workspace-mode accepts:

Mode	Behavior
bind-rw	Bind mount the host path read-write at /workspace (default).
bind-ro	Bind mount the host path read-only at /workspace.
copy-in-out	Copy the host path into a private staging directory, run against that staged tree, then sync changes back after exit.

Workspace mounts are nosuid,nodev,noexec by default and native Landlock denies
execution from /workspace. Use --workspace-exec for agent-mode workflows
that build and run test binaries from the workspace. Production mode rejects
writable executable workspaces; use an immutable --rootfs and explicit policy
files for production services.

Sandbox Home and Provider Config

Nucleus creates a private tmpfs home at /home/agent by default and sets the
workload HOME to that path. The home tmpfs is mounted nosuid,nodev,noexec
with mode 0700 and is owned by the configured workload uid/gid. Use
--home <container-path> to choose a different private home path; the path must
be absolute and must not overlap /workspace.

Provider CLIs that require config under $HOME should use explicit provider
config mounts instead of broad host bind mounts:

nucleus run \
  --home /home/agent \
  --provider-config-ro "$HOME/.aws:.aws" \
  --provider-config-ro "$HOME/.config/gcloud:.config/gcloud" \
  --provider-config-rw "$HOME/.config/gh:.config/gh" \
  -- ./agent

--provider-config-ro SOURCE:DEST and --provider-config-rw SOURCE:DEST are
repeatable. DEST may be absolute under the configured home, or relative to the
home directory. Read-only mounts are preferred for cloud credentials; read-write
mounts are intended only for tools that must refresh local tokens.

Agent Toolchain Rootfs

Mitos-style provider launchers can avoid depending on mutable host /bin,
/usr, /lib, or /nix binds by passing a pinned agent toolchain rootfs:

nucleus run \
  --service-mode strict-agent \
  --agent-toolchain-rootfs /nix/store/...-nucleus-agent-toolchain-rootfs \
  --workspace "$PWD" \
  --workspace-exec \
  -- claude

The dedicated flag is for agent, strict-agent, and mitos-agent modes. It
uses the same read-only rootfs mount path as --rootfs, but is rejected in
production mode so production services keep using --rootfs with attestation.

Build a rootfs with the Nix helper:

nucleus.lib.mkAgentToolchainRootfs {
  inherit pkgs;
  providerPackages = [
    # Derivations that provide claude/codex/gemini executables.
  ];
  extraPackages = [
    pkgs.rustc
    pkgs.cargo
  ];
}

The repository also exposes packages.${system}.agent-toolchain-rootfs as a
default shell/Git/compiler/package-manager rootfs. Integrations that need exact
provider CLIs should call mkAgentToolchainRootfs with pinned provider package
derivations and pass the resulting store path to --agent-toolchain-rootfs.

Image Snapshots

GPU Passthrough

Expose host GPUs to a container for CUDA, ROCm, or Mesa/Vulkan compute workloads.
GPU access is an explicit, audited privilege grant - Nucleus preserves its
full defense-in-depth stack (namespaces, capabilities, Landlock, cgroups)
while binding only the requested device nodes and relaxing only the ioctl
seccomp rule.

# Auto-detect and bind whatever GPU is present on the host
nucleus create --gpu auto --rootfs /nix/.../cuda-rootfs -- python train.py

# Pin a vendor and explicit devices (overrides discovery)
nucleus create --gpu nvidia \
  --gpu-device /dev/nvidia0 \
  --gpu-device /dev/nvidiactl \
  --rootfs /nix/.../cuda-rootfs -- nvidia-smi

# AMD ROCm
nucleus create --gpu amd --rootfs /nix/.../rocm-rootfs -- rocminfo

# Bind only the devices; the rootfs ships its own driver stack
nucleus create --gpu auto --no-gpu-driver-libs --rootfs /nix/.../rootfs -- ./workload

What happens when --gpu is set:

• Device nodes - the resolved /dev/nvidia*, /dev/dri/renderD*, /dev/kfd,
  /dev/nvidia-uvm*, etc. are bind-mounted into the container /dev at their
  host paths and chown'd to the workload identity.
• cgroup device allowlist - a BPF_PROG_TYPE_CGROUP_DEVICE program is
  attached to the container cgroup, allow-listing only the base /dev nodes
  plus the bound GPU devices (deny-by-default). This is best-effort: rootless
  launches and kernels without bpf(2) degrade to a warning, leaving the
  filesystem layer (only bound device nodes exist in /dev) as the gate.
• Seccomp - the restrictive terminal-only ioctl rule is replaced with an
  unconditional allow, since vendor driver ioctl request codes cannot be
  enumerated. All other seccomp restrictions remain.
• Driver support files - NVIDIA /proc/driver/nvidia, driver userspace
  library directories, and Vulkan/ICD/EGL manifests are bind-mounted read-only
  when present (--no-gpu-driver-libs skips this).
• Environment - NVIDIA_VISIBLE_DEVICES, NVIDIA_DRIVER_CAPABILITIES, and
  the EGL vendor manifest pointer are injected at exec time.
• gVisor - runsc receives OCI linux.devices entries (major/minor/type) so
  it creates the device nodes and installs the matching cgroup rules inside its
  sandbox, plus bind mounts for the driver support files.
• Auditing - the container_started event stream reports the vendor,
  visible devices, driver capabilities, and the relaxed-seccomp flag.

CLI flags:

Flag	Purpose
--gpu <auto|nvidia|amd|intel|all>	Enable GPU passthrough and select vendor(s)
--gpu-device <path>	Explicit device node (repeatable; overrides discovery)
--gpu-driver-capabilities <s>	NVIDIA_DRIVER_CAPABILITIES (default compute,utility)
--gpu-visible-devices <s>	NVIDIA_VISIBLE_DEVICES (default all)
--no-gpu-driver-libs	Do not bind host driver userspace libraries

--gpu is rejected in --service-mode production (declare GPU needs through an
attested rootfs instead). The same configuration is available programmatically
via the gpu field of the launch config document (--config/--config-fd).

Image Snapshots

Nucleus images are local directories containing a manifest, rootfs attestation,
store path list, optional overlay diff, and a signature for runtime-committed
images. They are not OCI/Docker images and are not pushed to or pulled from a
registry.

# Start an overlay-backed native container from a Nix rootfs
nucleus create -d \
  --name worker \
  --runtime native \
  --trust-level trusted \
  --rootfs /nix/store/...-worker-rootfs \
  --rootfs-mode overlay \
  -- /bin/sh -c 'echo committed > /tmp/result; sleep 3600'

# Commit the overlay upperdir as a signed thin image
nucleus image commit worker -o ./worker.nucleus-image --freeze

# Verify/load and inspect the image
nucleus image load ./worker.nucleus-image
nucleus image inspect ./worker.nucleus-image

# Run the manifest command, or override it after --
nucleus image run ./worker.nucleus-image -- /bin/cat /tmp/result

nucleus image commit requires a container launched with --rootfs-mode overlay; overlay rootfs mode is currently native-runtime only and production
mode rejects it. Runtime-committed images are signed with a host-local HMAC key.
Use --image-key-file PATH or set NUCLEUS_IMAGE_HMAC_KEY_FILE to pin that key
path; otherwise Nucleus creates an owner-only key under /var/lib/nucleus for
root or the user's data directory for non-root runs. Use the same key path when
one uid commits an image and another uid, such as root, runs or inspects it.
Nix-built images from nucleus.lib.mkImage live in /nix/store and omit
image.sig because Nix store/substituter trust is the integrity root.

Runtime commits record the original command, explicit environment variables,
workdir, uid/gid, supplementary groups, and an overlay diff. The diff exporter
preserves uid/gid, mode bits, mtimes, symlinks, xattrs, POSIX ACL xattrs, and
Linux file-capability xattrs; commit fails if the caller cannot preserve
metadata. Local image signatures cover both content and preserved metadata.

nucleus image commit reads the container's live overlay upperdir. Use
--freeze to freeze the recorded cgroup while the diff is copied; without it,
concurrent writes can be captured in a torn state. Freezing requires the
container to have a recorded cgroup path. Runtime-committed images are a
development/CLI workflow: the NixOS production module consumes build-time
mkImage images only and intentionally rejects images with overlay diffs. Image
run also constrains base.rootfs_path to a canonical /nix/store rootfs before
mounting it as the overlay lowerdir.

Image v2 is a single-rootfs plus optional single-diff directory format. There is
no layer chain, image store, registry push/pull, or ls/rm/cp image management
surface; operators manage image directories directly or use Nix store paths for
declarative production deployments.

Detached Mode

Use -d/--detach to run a container in the background as a systemd transient service. The CLI prints the container ID and exits immediately; systemd supervises the container process.

# Run a container in the background
nucleus create -d --memory 512M -- /bin/sleep 3600
# prints: a1b2c3d4e5f6...

# All management commands work with detached containers
nucleus state                        # list running containers
nucleus logs <container>             # view stdout/stderr (from journald)
nucleus logs -f <container>          # follow logs
nucleus logs -n 50 <container>       # last 50 lines
nucleus attach <container>           # exec into it
nucleus stop <container>             # graceful SIGTERM → SIGKILL
nucleus kill <container>             # send signal

# Detach works with all create flags
nucleus create -d \
  --name my-service \
  --memory 1G --cpus 2 \
  --network bridge -p 8080:80 \
  -- ./my-server

# systemd unit is named nucleus-<id-prefix>
systemctl status nucleus-a1b2c3d4e5f6
journalctl -u nucleus-a1b2c3d4e5f6

The systemd transient service uses KillMode=mixed and TimeoutStopSec=30, so systemctl stop also works for graceful shutdown. The --collect flag ensures the unit is garbage-collected after the container exits.

Production Mode

Production mode enforces strict security invariants:

• Forbids --allow-degraded-security, --allow-chroot-fallback, and native --network host
• Permits --allow-host-network only with --network gvisor-host --runtime gvisor
• Requires explicit --memory limit
• Requires successful cgroup creation (no fallback to running without limits)
• Egress policy failures are fatal where Nucleus owns the network namespace; gvisor-host cannot use Nucleus egress policy
• Bridge DNS must be configured explicitly (no public resolver defaults)

# Run a long-running service with production hardening
nucleus run \
  --service-mode production \
  --trust-level trusted \
  --memory 1G --cpus 2 --pids 256 \
  --rootfs /nix/store/...-my-service-rootfs \
  --verify-rootfs-attestation \
  --require-kernel-lockdown integrity \
  --network bridge --dns 10.0.0.1 \
  --egress-allow 10.0.0.0/8 \
  --egress-domain api.example.com \
  --egress-tcp-port 443 --egress-tcp-port 8443 \
  --health-cmd "curl -sf http://localhost:8080/health" \
  --health-interval 30 --health-retries 3 \
  --secret /run/secrets/tls-cert:/etc/tls/cert.pem \
  --systemd-credential db-url:/run/secrets/db-url \
  --volume /var/lib/myservice:/var/lib/myservice:rw \
  -e CONFIG_PATH=/etc/myservice/config.toml \
  --sd-notify \
  -p 127.0.0.1:8080:8080 \
  -- /bin/my-service --config /etc/myservice/config.toml

# gVisor with network access (sandbox network stack)
nucleus run \
  --service-mode production \
  --runtime gvisor \
  --gvisor-platform kvm \
  --memory 512M \
  --network bridge --dns 10.0.0.1 \
  --rootfs /nix/store/...-proxy-rootfs \
  -- /bin/proxy

Strict Agent Mode

Strict agent mode (--service-mode strict-agent, --service-mode mitos-agent, or --strict-agent) keeps agent-style execution while making isolation setup fail closed:

• Forbids --allow-degraded-security, --allow-chroot-fallback, and native --network host
• Permits --allow-host-network only with --network gvisor-host --runtime gvisor
• Requires successful cgroup creation and successful application of configured limits
• Requires pivot_root in native mode; no chroot fallback
• Requires seccomp enforcement; --seccomp-mode trace is rejected
• Requires Landlock enforcement for native runtime
• Requires user namespace UID/GID mapping when running as host root or rootless
• Keeps network mode none by default; bridge mode requires explicit --dns

Strict agent mode does not require a production Nix rootfs, rootfs attestation, health checks, readiness probes, sd_notify, systemd transient services, or NixOS module deployment.

# Run an ephemeral agent with fail-closed native isolation
nucleus run \
  --service-mode strict-agent \
  --runtime native \
  --trust-level trusted \
  --memory 1G --cpus 2 \
  --context ./ctx \
  -- ./agent

Security Policy Files

Nix defines the service and the root filesystem; separate files define security policy (what the process is allowed to do at the kernel level). This separation keeps deployments declarative, security config auditable, and runtime inputs reproducible without coupling policy changes to application rebuilds.

# Run with external security policies
nucleus run \
  --service-mode production \
  --rootfs /nix/store/...-my-service-rootfs \
  --memory 512M --cpus 1 \
  --seccomp-profile ./config/my-service.seccomp.json \
  --seccomp-profile-sha256 abc123... \
  --caps-policy ./config/my-service.caps.toml \
  --landlock-policy ./config/my-service.landlock.toml \
  -- /bin/my-service

Seccomp profile (JSON - OCI-native format, tooling emits it directly):

{
  "defaultAction": "SCMP_ACT_KILL_PROCESS",
  "architectures": ["SCMP_ARCH_X86_64"],
  "syscalls": [
    {
      "names": ["read", "write", "close", "openat", "fstat",
                "mmap", "munmap", "brk", "futex", "clock_gettime"],
      "action": "SCMP_ACT_ALLOW"
    }
  ]
}

Capability policy (TOML):

# config/my-service.caps.toml
[bounding]
keep = []          # empty = drop all

[ambient]
keep = []

Landlock policy (TOML):

# config/my-service.landlock.toml
min_abi = 3

[[rules]]
path = "/bin"
access = ["read", "execute"]

[[rules]]
path = "/etc/myservice"
access = ["read"]

[[rules]]
path = "/run/secrets"
access = ["read"]

[[rules]]
path = "/tmp"
access = ["read", "write", "create", "remove"]

Seccomp Profile Generation

Profiles shouldn't be hand-written from scratch. Use trace mode to record actual syscall usage, then generate a minimal profile:

# 1. Run in trace mode – all syscalls allowed but logged
nucleus run \
  --seccomp-mode trace \
  --seccomp-log ./trace.ndjson \
  --rootfs /nix/store/...-my-service-rootfs \
  --memory 512M \
  -- /bin/my-service

# 2. Generate minimal profile from trace
nucleus seccomp generate ./trace.ndjson -o config/my-service.seccomp.json

# 3. Review and tighten (remove anything surprising)
# 4. Commit – Nix pins the SHA-256 hash
# 5. Run in enforce mode
nucleus run \
  --seccomp-profile ./config/my-service.seccomp.json \
  --seccomp-profile-sha256 "$(sha256sum config/my-service.seccomp.json | cut -d' ' -f1)" \
  -- /bin/my-service

Trace mode requires root or CAP_SYSLOG (reads /dev/kmsg). It is rejected in production mode - it is a development tool only.

Multi-Container Topologies

Nucleus includes a Compose-equivalent for managing multi-container stacks using TOML configuration with dependency ordering.

# topology.toml
name = "myapp"

[networks.internal]
subnet = "10.42.0.0/24"

[volumes.db-data]
volume_type = "persistent"
path = "/var/lib/nucleus/myapp/db"
owner = "70:70"

[volumes.cache]
volume_type = "ephemeral"
size = "128M"

[services.postgres]
rootfs = "/nix/store/...-postgres"
command = ["postgres", "-D", "/var/lib/postgresql/data"]
memory = "2G"
cpus = 2.0
networks = ["internal"]
volumes = [
  "db-data:/var/lib/postgresql/data",
  "cache:/var/cache/postgresql"
]
health_check = "pg_isready -U myapp"

[services.web]
rootfs = "/nix/store/...-web"
command = ["/bin/web-server"]
memory = "512M"
networks = ["internal"]
nat_backend = "userspace"
port_forwards = ["8443:8443"]
egress_allow = ["10.42.0.0/24"]
egress_domains = ["api.example.com"]

[[services.web.depends_on]]
service = "postgres"
condition = "healthy"

# Validate topology and show dependency order
nucleus compose validate -f topology.toml

# Bring up all services in dependency order
nucleus compose up -f topology.toml

# Show service status
nucleus compose ps -f topology.toml

# Tear down in reverse dependency order
nucleus compose down -f topology.toml

Container Management

# List running containers
nucleus ps

# List all containers (including stopped)
nucleus ps --all

# Show resource usage statistics
nucleus stats

# View logs for a detached container (from systemd journal)
nucleus logs <container>
nucleus logs -f <container>          # follow output
nucleus logs -n 100 <container>      # last 100 lines

# Stop a container (SIGTERM, then SIGKILL after timeout)
nucleus stop <container>
nucleus stop --timeout 30 <container>

# Kill a container with a specific signal
nucleus kill <container>
nucleus kill --signal TERM <container>

# Remove a stopped container
nucleus rm <container>
nucleus rm --force <container>

# Attach to a running container
nucleus attach <container>
nucleus attach <container> -- /bin/bash

# Checkpoint a running container (requires root, CRIU)
nucleus checkpoint <container> --output /path/to/checkpoint

# Restore from checkpoint
nucleus restore --input /path/to/checkpoint

NixOS Module

Nucleus provides a declarative NixOS module for running containers as systemd services. Each container is managed as a nucleus-<name>.service unit with journald logging, sd_notify readiness, and automatic restart.

Flake Setup

{
  inputs.nucleus.url = "github:wiggum-cc/nucleus";

  outputs = { self, nixpkgs, nucleus, ... }: {
    nixosConfigurations.myhost = nixpkgs.lib.nixosSystem {
      system = "x86_64-linux";
      modules = [
        nucleus.nixosModules.default
        ./configuration.nix
      ];
    };
  };
}

Service Configuration

{ pkgs, nucleus, ... }:

let
  # Build a minimal rootfs containing only the packages your service needs.
  # This replaces host bind mounts with a locked-down Nix closure.
  proxyRootfs = nucleus.lib.mkRootfs {
    inherit pkgs;
    packages = [ my-proxy-pkg pkgs.cacert pkgs.curl ];
  };
in
{
  services.nucleus = {
    enable = true;
    package = nucleus.packages.x86_64-linux.default;

    containers.sigid-proxy = {
      enable = true;
      command = [ "/bin/sigid-proxy" "--config" "/etc/sigid/proxy.toml" ];
      rootfs = proxyRootfs;
      user = "sigid-proxy";
      group = "sigid-proxy";

      # Resource limits (required in production mode)
      memory = "1G";
      cpus = 2.0;
      pids = 256;

      # Security policy files (separate from Nix, auditable by security engineers)
      seccompProfile = {
        path = ./config/sigid-proxy.seccomp.json;
        sha256 = "abc123...";  # Nix verifies at build time
      };
      capsPolicy = ./config/sigid-proxy.caps.toml;
      landlockPolicy = ./config/sigid-proxy.landlock.toml;

      # Optional hardening toggles
      verifyRootfsAttestation = true;
      seccompLogDenied = true;
      requireKernelLockdown = "integrity";

      # Networking
      network = "bridge";
      natBackend = "auto";  # or "userspace" to force slirp4netns
      dns = [ "10.0.0.1" ];  # internal resolver – no public DNS default
      portForwards = [ "127.0.0.1:8080:8080" "127.0.0.1:8443:8443" ];

      # Egress policy – audited outbound access
      egressAllow = [ "10.0.0.0/8" ];
      egressDomains = [ "api.example.com" ];
      egressTcpPorts = [ 443 8443 ];

      # Credential broker alternative for bearer-token APIs.
      # Mutually exclusive with egressAllow / egressDomains above.
      # credentialBroker = "10.0.42.1:8080";
      # credentialBrokerNoProxyEnv = false;

      # Health checking
      healthCheck = "curl -sf http://localhost:8080/health";
      healthInterval = 30;
      healthRetries = 3;
      healthStartPeriod = 10;

      # Secrets (mounted read-only)
      secrets = [
        { source = config.age.secrets.proxy-tls.path; dest = "/etc/tls/cert.pem"; }
      ];

      # systemd-creds integration
      credentials = [
        {
          name = "proxy-key";
          source = config.age.secrets.proxy-key.path;
          dest = "/run/secrets/proxy-key";
          encrypted = false;
        }
      ];

      # Volumes (bind-mounted host paths)
      volumes = [
        {
          source = "/var/lib/sigid-proxy";
          dest = "/var/lib/sigid-proxy";
          createHostPath = true;
        }
      ];

      # Environment
      environment = {
        RUST_LOG = "info";
        CONFIG_PATH = "/etc/sigid/proxy.toml";
      };

      # systemd integration
      sdNotify = true;  # Type=notify, passes NOTIFY_SOCKET into container
    };
  };
}

Writable bind volumes are automatically added to the generated systemd unit's ReadWritePaths. When createHostPath = true, the NixOS module creates the host directory with systemd-tmpfiles before the container starts. If the container declares a workload user/group, those become the default tmpfiles owner for new writable paths unless the volume overrides them.

Credentials declared via credentials = [ ... ] use systemd's credential pipeline (LoadCredential or LoadCredentialEncrypted) and are mounted into the container through Nucleus's secret path. The CLI flag --systemd-credential NAME:DEST resolves NAME from CREDENTIALS_DIRECTORY at runtime.

For bearer-token API clients, the NixOS module exposes credentialBroker = "IP:PORT"; and credentialBrokerNoProxyEnv = true;. This maps to --credential-broker and installs broker-only egress, so leave egressAllow, egressDomains, and egress port allowlists empty when using it.

Set image = appImage; instead of rootfs = proxyRootfs; when a service should
consume a Nix-built image produced by nucleus.lib.mkImage. rootfs and
image are mutually exclusive. When command = [ ];, the module uses the image
manifest command. The NixOS production launcher currently supports build-time
images without overlay diffs; committed runtime image diffs are a local CLI
workflow.

Set user, group, and optional supplementaryGroups on a NixOS container definition when the workload should run as a dedicated service account instead of root.

Topology Services

Topologies can also be managed as systemd services:

{
  services.nucleus = {
    enable = true;
    package = nucleus.packages.x86_64-linux.default;

    topologies.myapp = {
      enable = true;
      configFile = ./topology.toml;
    };
  };
}

This creates a nucleus-topology-myapp.service (Type=oneshot, RemainAfterExit) that runs nucleus compose up on start and nucleus compose down on stop.

What the Module Generates

For each enabled container, the module creates a systemd service:

• Unit: nucleus-<name>.service, ordered after network-online.target
• Type: notify (when sdNotify = true) or simple
• Restart: on-failure with 5s backoff
• Logging: stdout/stderr captured to journald with SyslogIdentifier=nucleus-<name>
• Command: nucleus run --service-mode production ... with all configured options
• Workload identity: Nucleus itself starts as root for setup, then drops the container workload to the configured user / group before exec
• Hardening: ProtectSystem=strict, ProtectHome=true at the systemd level (defense-in-depth)

Building a Rootfs

Use nucleus.lib.mkRootfs to build a minimal, reproducible root filesystem:

nucleus.lib.mkRootfs {
  inherit pkgs;
  name = "my-service-rootfs";  # optional, defaults to "nucleus-rootfs"
  packages = [
    my-service-package
    pkgs.cacert       # TLS certificates
    pkgs.curl         # for health checks
    pkgs.busybox      # minimal coreutils
  ];
}

This produces a Nix store path containing /bin, /lib, /etc, etc. from the specified packages. It is mounted read-only inside the container, replacing the host bind mounts used in agent mode.

mkRootfs also emits a .nucleus-rootfs-sha256 manifest at the root of the closure. Use --verify-rootfs-attestation or verifyRootfsAttestation = true; to require that manifest to match the mounted rootfs at startup.

For ephemeral provider agents, use nucleus.lib.mkAgentToolchainRootfs
instead. It layers a broad agent development toolchain on top of mkRootfs,
keeps /bin/sh and /usr/bin/env compatibility paths available, and accepts
provider CLI packages through providerPackages.

Building an Image

Use nucleus.lib.mkImage to package a Nix rootfs plus default process config as
a reproducible Nucleus image:

let
  appRootfs = nucleus.lib.mkRootfs {
    inherit pkgs;
    name = "my-service-rootfs";
    packages = [
      my-service-package
      pkgs.cacert
      pkgs.curl
    ];
  };

  appImage = nucleus.lib.mkImage {
    inherit pkgs;
    name = "my-service-image";
    rootfs = appRootfs;
    config = {
      command = [ "/bin/my-service" "--config" "/etc/my-service.toml" ];
      env = {
        RUST_LOG = "info";
      };
      workdir = "/";
      uid = 0;
      gid = 0;
    };
  };
in
{
  services.nucleus.containers.my-service = {
    enable = true;
    image = appImage;
    command = [ ]; # use the image manifest command
    memory = "512M";
    cpus = 1.0;
  };
}

mkImage writes manifest.json, rootfs.sha256, and store-paths into a Nix
store output. Build-time images are cold and thin: the rootfs remains a Nix
store path, and the image manifest has no overlay diff unless it was produced by
the CLI nucleus image commit workflow.

Security Notes

Do not pass secrets via -e / --env. Environment variables are visible in /proc/<pid>/environ to any process that can read it (mitigated by hidepid=2 in production mode, but not in agent mode). Use --secret instead when a file works. If a provider CLI requires sensitive environment variables, use --env-fd FD; the fd carries a JSON object such as {"OPENAI_API_KEY":"..."} or a JSON array of KEY=VALUE strings so the values are not exposed through Nucleus argv.

Prefer credential brokers for bearer-token APIs. If untrusted code can drive a provider CLI, do not place the bearer token in the sandbox environment. Run a host-side broker that holds the credential, injects it into approved upstream requests, rate-limits and audits usage, and start Nucleus with --credential-broker IP:PORT so the sandbox can only reach that broker endpoint. Proxy environment variables are client hints, not the security boundary: Nucleus enforces the boundary with namespace-local egress rules, and non-HTTP clients that ignore proxy variables are denied unless they speak to the broker endpoint.

Protect the local image signing key. Runtime-committed image directories are verified with the host-local HMAC key selected by NUCLEUS_IMAGE_HMAC_KEY_FILE or the default owner-only key path. Treat that file like deployment signing material: do not share it across trust domains unless those hosts should be able to trust and produce each other's local image snapshots.

Privilege dropping is explicit. Nucleus must start with elevated privileges to create namespaces, mount filesystems, and configure cgroups/networking. Use --user / --group (or the NixOS module's user / group options) so the workload itself does not continue running as root after setup. In production mode, staged secrets under /run/secrets are re-owned to that workload identity.

Agent mode is not hardened. By design, agent mode applies several security mechanisms on a best-effort basis: seccomp and Landlock failures are warn-and-continue (with --allow-degraded-security), chroot fallback is available (with --allow-chroot-fallback), bridge DNS defaults to public resolvers (8.8.8.8), and cgroup creation failures are non-fatal. Operators requiring strict isolation for ephemeral workloads should use --service-mode strict-agent; operators deploying long-running NixOS services should use production mode.

Service Modes

Feature	Agent Mode	Strict Agent Mode	Production Mode
Service mode	--service-mode agent (default)	--service-mode strict-agent (alias: --service-mode mitos-agent)	--service-mode production
Degraded security	Allowed with flag	Forbidden	Forbidden
Chroot fallback	Allowed with flag	Forbidden	Forbidden
Host networking	Allowed with flag	Native host forbidden; gvisor-host allowed with gVisor + explicit opt-in	Native host forbidden; gvisor-host allowed with gVisor + explicit opt-in
Cgroup limits	Best-effort	Required (fatal on create/apply failure)	Required (fatal on create/apply failure)
Bridge DNS	Defaults to 8.8.8.8/8.8.4.4	Must be configured explicitly unless credential broker mode disables DNS	Must be configured explicitly unless credential broker mode disables DNS
Rootfs	Host bind mounts unless --rootfs (optionally with --rootfs-mode overlay) or --agent-toolchain-rootfs is supplied	Host bind mounts unless --rootfs (optionally with --rootfs-mode overlay) or --agent-toolchain-rootfs is supplied	Pre-built Nix closure (--rootfs) or build-time mkImage image without an overlay diff
Workspace	Optional /workspace; bind/copy-in-out for agents	Optional /workspace; bind/copy-in-out for agents	Optional, non-executable unless read-only or policy-specific
Egress policy	Optional	Optional	Deny-all default where enforceable; unavailable with gvisor-host
Memory limit	Optional	Optional	Required
PID 1 init	Direct exec	Direct exec	Mini-init with zombie reaping + signal forwarding
Workload uid/gid	Root by default	User namespace remapping required when running as host root	Configurable post-setup drop via --user / --group
Secrets	In-memory tmpfs	In-memory tmpfs	In-memory tmpfs with volatile zeroing
/proc	Mounted normally	Mounted normally	hidepid=2 (hides other processes)
Mount audit	Skipped	Skipped	Post-setup flag verification (fatal)
Seccomp trace mode	Allowed	Forbidden	Forbidden
Landlock ABI	Best-effort	Full enforcement required on native	V3 minimum required
Health checks	Optional	Optional	Optional
sd_notify	Optional	Optional	Optional
Security policies	Optional	Optional	Optional (recommended)

Overlay rootfs mode is a writable development snapshot mode, not the strict
production posture. To support overlayfs copy-up, Nucleus retains
CAP_DAC_OVERRIDE and CAP_FOWNER in the workload and grants native Landlock
read/write/execute access to /. Use bind rootfs mode for the default-deny
Landlock and all-capabilities-dropped posture.

Egress Policy

When production bridge mode runs without --egress-allow or --egress-domain, Nucleus installs a strict deny-all OUTPUT policy, including DNS.
When --egress-allow or --egress-domain is specified, Nucleus applies iptables OUTPUT chain rules inside the container's network namespace:

1. Allow loopback traffic
2. Allow established/related connections
3. Allow DNS to configured resolvers
4. Resolve permitted domains to IPv4 /32 rules at startup
5. Allow traffic to permitted CIDRs and resolved domain addresses (optionally restricted to specific ports)
6. Log denied packets (rate-limited, nucleus-egress-denied: prefix)
7. Drop everything else

# Allow outbound to internal network on HTTPS only
nucleus run --network bridge --dns 10.0.0.1 \
  --egress-allow 10.0.0.0/8 --egress-tcp-port 443 \
  -- ./my-service

# Allow outbound to a provider API domain on HTTPS only
nucleus run --network bridge --dns 10.0.0.1 \
  --egress-domain api.example.com --egress-tcp-port 443 \
  -- ./provider-client

# Production deny-all egress, including DNS
nucleus run --service-mode production --network bridge --dns 10.0.0.1 \
  -- ./isolated-service

Domain egress entries are exact DNS names, not wildcard or suffix rules. Nucleus resolves each domain with the supervisor host resolver before installing the namespace-local iptables policy, keeps only IPv4 answers, and fails startup if a domain has no IPv4 address. Long-running services that depend on provider IP rotation should restart after DNS changes, use provider-published CIDR ranges, or route traffic through a stable internal proxy and allow that proxy CIDR instead.

Credential Broker Egress

--credential-broker IP:PORT is the first-class Nucleus path for bearer-token API clients that must run inside an untrusted sandbox. The actual broker process is host-side and outside Nucleus: it owns the real secret, authenticates outbound requests, enforces upstream method/path/destination limits, and writes the audit log. Nucleus enforces the sandbox side by installing a deny-by-default policy that allows only TCP to the broker /32 and disables DNS from the sandbox.

## Broker listens on the host side of the bridge, for example 10.0.42.1:8080.
## Nucleus injects HTTP_PROXY/H...