Pp Loops

<!-- GENERATED FILE — DO NOT EDIT.
     This file is a verbatim mirror of library/marketing/loops/SKILL.md,
     regenerated post-merge by tools/generate-skills/. Hand-edits here are
     silently overwritten on the next regen. Edit the library/ source instead.
     See the repository agent guide, section "Generated artifacts: registry.json, cli-skills/". -->

Loops CLI agent guide

Inspect COMMANDS.md for every command's effect, then call the specific command's --help. Use the installed binary or go run ./cmd/loops-pp-cli from a source checkout.

Prerequisites: Install the CLI

This skill drives the loops-pp-cli binary. You must verify the CLI is installed before invoking any command from this skill. If it is missing, install it first:

1. Install via the Printing Press installer. It defaults binaries to $HOME/.local/bin on macOS/Linux and %LOCALAPPDATA%\Programs\PrintingPress\bin on Windows:

   npx -y @mvanhorn/printing-press-library install loops --cli-only

2. Verify: loops-pp-cli --version
3. Ensure the reported install directory is on $PATH for the agent/runtime that will invoke this skill.

If the npx install fails (no Node, offline, etc.), fall back to a direct Go install (requires Go 1.26.6 or newer). This installs into $GOPATH/bin (default $HOME/go/bin), so add that directory to $PATH instead:

go install github.com/mvanhorn/printing-press-library/library/marketing/loops/cmd/loops-pp-cli@latest

If --version reports "command not found" after install, the runtime cannot see the binary directory on $PATH. Do not proceed with skill commands until verification succeeds.

This CLI covers all 68 operations in the Loops API and adds agent-safe team verification, lifecycle inventory, and campaign preflight checks. It previews writes with private values redacted and requires explicit execution for sends and changes.

Safe sequence

1. Run loops-pp-cli --version and loops-pp-cli doctor --dry-run --json. The dry doctor makes no network request and sends no email.
2. Select a Loops credential through LOOPS_API_KEY or private local credential storage. Keep different teams' credentials and profiles separate. For account-specific work, run team verify --team 'Expected Team' --agent before reads; read-only checks may use LOOPS_EXPECTED_TEAM instead. Every mutation verifies the exact API team again and still requires an explicit --team flag.
3. Use --agent --no-learn --no-cache and --select for minimal JSON output and local retention. contacts find --agent omits identity fields by default. Do not place contact values in teach/recall queries, logs, receipts, source, fixtures, or public descriptions.
4. Inspect a mutation's default redacted preview. Use --execute --team 'Expected Team' only after reviewing the target and effect. Sends also need --confirm-send and an explicit --idempotency-key; campaign publication, suppression removal, and deletes need their respective confirmation flags. --yes does not bypass these gates.
5. After any uncertain send result, inspect Loops before taking further action. The local journal blocks reusing that team/key automatically. A definite 429 can be retried later with the same key. Never create a fresh key merely to bypass a duplicate warning.