Loom onboarding

Skill joycastle/loom/skills/loom-onboarding

Guide a new or returning user through loom setup, configuration, first sync, history intake, and the code-vs-data boundary. Use when the user asks to initialize loom, follow ONBOARDING.md, configure sources, collect historical work traces, or explain what lives in the repo versus ~/.loom.From its SKILL.md

Install

npx -y skills add joycastle/loom --skill loom-onboarding

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

13 stars13 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

2.7 KB, as published. Nobody here has run it

Loom Onboarding

Workflow

Use ONBOARDING.md as the source-of-truth runbook. Work one phase at a time, inspect command output before proceeding, and do not skip the privacy boundaries.

If loom is not on PATH, run commands from the repo root with python3 -m loom.

Guardrails

Confirm before irreversible or external actions: deleting or moving files, --push, creating a remote repo, sharing data, or writing cron/automation.
Put credentials only in ~/.loom/.env with mode 600. Never paste secrets into repo files, markdown notes, logs, or chat output.
Explain the split early: this repo is tool code; ~/.loom/ is the private instance containing config.json, .env, data/, and vault/.
Treat loom sync as local collection/render/vault commit. Treat loom sync --push as external backup and require confirmation.

Phase A: Environment

Run:

loom --help || python3 -m loom --help
python3 -m pytest tests/ -q

Ask for the user's name, git author emails or names, and the root directory to scan for git repos. Default the scan root to ~/Documents if the user has no preference.

Run:

loom init

Let the user enter secrets interactively when Feishu is configured.

Phase B: First Sync

Run without push:

loom sync
loom today
loom search <keyword>

Explain the result in concrete terms: entries live in ~/.loom/data/entries.jsonl, search index and journals are derived, and collection is read-only except for writing the loom instance.

Phase C: Intake

Use explicit commands for loose material:

loom doc add <paths...> --to <category> --tags <tags>
loom data add <csv-or-xlsx...> --to <topic> --from <upstream> --code <code-files>
loom note "<text>" --to <category> --tags <tags>
loom report import <daily-report.xlsx>

Do not bulk-import raw data by directory. CSV/XLSX files must go through loom data add so raw files remain local under _data/ and only data cards become searchable.

Phase D: Backup

Only after the user confirms private backup, verify the vault remote and risk with $loom-vault-ops. Then run:

loom sync --push

Do not set cron or recurring automation without explicit confirmation.

Phase E: Next Skills

For topic classification, switch to $loom-topic-triage. For daily reports, use $loom-daily-report. For session summaries, use $loom-session-digest.