p4n4-docs

CLI Reference

Install: pip install p4n4

Global options

p4n4 [--version] [--help]

`p4n4 init NAME`

Interactive project wizard. Scaffolds stack files, generates secrets, writes .p4n4.json.

p4n4 init my-project                       # IoT layer (default)
p4n4 init my-project --layer iot,ai        # multiple layers
p4n4 init my-project --layer all           # iot + ai + edge
p4n4 init my-project --no-interactive      # skip wizard, use defaults
p4n4 init my-project --source-iot ../p4n4-iot   # scaffold from a local checkout (offline)

The wizard prompts for the InfluxDB org, timezone and service passwords (blank means generate one), including the Node-RED editor password, which is written with NODE_RED_USER=admin to the IoT .env. The edge layer gets the compose file, runner, model directories and .env, sharing the InfluxDB token, org and timezone with the other layers. Unknown layer names are rejected.

Flag	Description
`--layer <names>`	Layers to enable: `iot`, `ai`, `edge`, `all`, or comma-separated
`--no-interactive`	Skip the wizard and auto-generate all secrets
`--source-iot PATH` / `--source-ai PATH` / `--source-edge PATH`	Use a local stack checkout instead of cloning

Layout: single-layer projects place docker-compose.yml, config/, scripts/, and .env at the project root. Multi-layer projects give each layer its own subdirectory (<project>/iot/, <project>/ai/, <project>/edge/) so the stacks run as separate Compose projects; shared .env keys (e.g. INFLUXDB_TOKEN) are written identically to every layer. See Getting Started.

`p4n4 add STACK`

Add a stack to an existing project.

p4n4 add ai
p4n4 add edge --path /srv/my-project

`p4n4 remove STACK`

Remove a stack from an existing project.

p4n4 remove edge
p4n4 remove ai --yes       # skip confirmation

`p4n4 up [STACK]`

Start one or all enabled stacks in dependency order (iot → ai → edge).

p4n4 up                    # all stacks
p4n4 up iot                # IoT stack only
p4n4 up --no-detach        # foreground mode

`p4n4 down [STACK]`

Stop one or all stacks, in reverse dependency order (dependents stop before iot removes the shared network).

p4n4 down
p4n4 down ai
p4n4 down --volumes        # also remove volumes

`p4n4 status`

Show container status for all enabled stacks. Multi-layer projects print one table per stack.

`p4n4 logs [SERVICE]`

Tail logs from services.

p4n4 logs                          # single-layer project: all services
p4n4 logs grafana --tail 50        # one service
p4n4 logs --stack ai               # multi-layer project: pick a stack
p4n4 logs --no-follow              # multi-layer: dump all stacks and exit

In multi-layer projects, pass --stack <name> to follow one stack’s logs, or --no-follow to print logs from every stack once.

`p4n4 validate`

Validate .p4n4.json, required stack files, and .env keys (the IoT layer requires NODE_RED_USER and NODE_RED_PASSWORD). Multi-layer projects are checked per layer directory (iot/…, ai/…, edge/…).

`p4n4 upgrade [STACK]`

Pull latest Docker images.

p4n4 upgrade               # all stacks
p4n4 upgrade iot

`p4n4 secret ACTION`

Action	Description
`show`	Show masked secrets from `.env` (per stack in multi-layer projects)
`rotate`	Re-generate passwords and tokens (see below)
`generate`	Print new secrets to stdout

rotate replaces whichever of these keys are present:

IoT: INFLUXDB_PASSWORD, INFLUXDB_TOKEN, GRAFANA_PASSWORD, NODE_RED_PASSWORD
AI: LETTA_SERVER_PASSWORD, N8N_BASIC_AUTH_PASSWORD, N8N_ENCRYPTION_KEY

In multi-layer projects, rotate updates every layer’s .env and writes the same new value to keys shared across stacks (e.g. INFLUXDB_TOKEN in both iot/.env and ai/.env), so cross-stack credentials never drift.

`p4n4 ei`

Not implemented in v0.1: deploy, run and status print “not yet implemented” and exit 1. Manage the runner with the edge stack’s make deploy-model / make up and its HTTP API instead.

Subcommand	Planned behavior
`deploy MODEL`	Copy a `.eim` or `.onnx` model to its models directory and recreate the runner
`run`	Start the inference runner
`status`	Show runner container status

list, infer, update and info are specified for v0.2 (F-0.2.4).

`p4n4 template`

Subcommand	Description
`search [QUERY]`	List templates from the community registry
`install NAME`	Install a template into the current project
`list`	Show installed templates