Pp Wanderlog

<!-- GENERATED FILE — DO NOT EDIT.
     This file is a verbatim mirror of library/travel/wanderlog/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/". -->

Wanderlog trip planning

Use wanderlog-pp-cli for creating and maintaining Wanderlog trips. It records itinerary and reservation data; it does not book or pay for travel.

Prerequisites: Install the CLI

This skill drives the wanderlog-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 wanderlog --cli-only

2. Verify: wanderlog-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/travel/wanderlog/cmd/wanderlog-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.

Public reads — guides, geos, places, category lists, and shared plans — need no cookie; writes do. With WANDERLOG_COOKIE, plan clone/plan fill and the fine-grained plan editor write through ShareDB, with previews before --apply. REST writes such as trips create use explicit --dry-run; removing that flag executes the write.

Authentication

Public plan/place reads work without a cookie. Account reads and writes require configured WANDERLOG_COOKIE. Use auth status --agent to inspect configuration without printing credentials. The user runs auth setup in their own terminal; never ask for a cookie in chat. There is no auth login. A configured cookie is not proof it is still valid.

Start with the task

Use the commands below directly when their syntax is sufficient. For missing workflow schemas, use agent-context --task review (or create / edit) with --agent; avoid loading the full inventory. For other tasks, use which "<task>" --agent, then the matching command's --help. agent-context defaults to a schema-4 summary with the complete command inventory; use --command "plan day" for one flag schema. --full restores legacy schema 3 and is expensive. --agent selects machine-readable output; it does not make a write a preview.

Creation

Resolve a destination with geos autocomplete --help; do not guess a geo ID. Preview the exact title, destinations, dates and privacy:

wanderlog-pp-cli trips create --geo-ids '[86696]' --title 'Trip draft' --start-date 2026-10-05 --end-date 2026-10-07 --privacy private --dry-run --agent

trips create uses the REST API: removing --dry-run creates the trip immediately. It has no --apply flag. Validate the target and user intent before running the real creation. It rejects invalid destination IDs, date order and privacy. A smoke probe can report validation: skipped; a request preview is not a successful creation.

Token-efficient planning

Choose the first read by scope. For a complete review of known days, read plan days --target-key KEY --days 1-N --travel-mode walking --agent once (replace N and the mode with the actual trip values). This includes full notes, schedules, coordinates, shared bookings, saved travel and checks; inspect every selected day and global constraint. For one day use plan day. If the day range is unknown, use a slim plan outline to find it.

For orientation or a focused question on a large trip, use plan overview, then only the relevant plan days or plan block get --block-ids. Overview omits ordinary stop details: follow detail references before judging their constraints. A complete review must read every day's full notes, even on a long trip; use consecutive bounded day groups if needed. Avoid an overview followed by all full days when a complete read already serves the task.

Reuse travel estimates, warnings and shared constraints already returned. Request plan route legs, plan inspect or an overview only for missing information (such as a cross-day check), changed data or lost context. Missing saved travel remains unknown. Markdown and raw text are optional duplicate representations.

For iterative day planning, use --save-state FILE, then --since FILE --save-state FILE. Only merge into the matching baseline still present in model context; after a new session or context loss, read a full day again. Unchanged delta fields inherit; explicit empty values clear fields. Preserve deletion, order, warnings and unknown travel coverage. A full fallback replaces the baseline. State files are private local content, not model memory or server cursors; delta reads still fetch live API data.

Batch additions with plan block add-batch and existing-block changes with plan edit, then verify affected days after applying; add an overview only when changes need cross-day checks. Preview the entire batch before applying. Do not drop notes or constraints to shrink output. See token efficiency for contracts and measurement limits, and batch creation for input examples.