site
/site
Pracuje z serwisem Archival w bieżącym folderze: dodaje strony i treści, naprawia build, wgrywa pliki i wdraża serwis.
- Autor
- archival-dev ↗
- Repozytorium
- archival ↗
- Licencja
- AGPL-3.0
- Ostatnia zmiana repozytorium
- 12.10.2026
Ocena i bezpieczeństwo
- Pliki w paczce
- 1
- Skrypty
- 0
- Komentarze HTML
- 0
- Adresy URL w SKILL.md
- 2
Brak sygnałów. Repozytorium zmienione 12.10.2026, bez odwołań do wycofanych modeli i API.
AGPL-3.0, copyleft: przeróbki udostępniane innym muszą mieć tę samą licencję.
Co robi i kiedy
Pracuje z serwisem Archival w bieżącym folderze: dodaje strony i treści, naprawia build, wgrywa pliki i wdraża serwis.
Oryginalny opis (EN)
Work on the Archival site in the current folder - add pages or kinds of content, add or edit content, fix a failing archival build, upload images and files, connect the folder to a hosted archival.dev site, and deploy. Use when the folder contains archival_objects.toml (or a legacy objects.toml), when someone asks to change, fix, publish or deploy their Archival site, or when a session-start note says this folder is an Archival site.
Jak działa
- 1. Look first
- 2. Baseline
- 3. Make the change the Archival way
- 4. Verify
- 5. Media
- 6. Commit and deploy
Podgląd SKILL.md
---
name: site
description: Work on the Archival site in the current folder - add pages or kinds of content, add or edit content, fix a failing archival build, upload images and files, connect the folder to a hosted archival.dev site, and deploy. Use when the folder contains archival_objects.toml (or a legacy objects.toml), when someone asks to change, fix, publish or deploy their Archival site, or when a session-start note says this folder is an Archival site.
---
# Work on the Archival site in this folder
This folder is the site. You have a shell, the `archival` CLI builds here, and
the hosted platform is reached through git and `archival upload`. The person
may be a developer or may not; either way, every change keeps the site one they
can go on editing in the Archival editor.
The shared references live beside the `new` skill:
`${CLAUDE_PLUGIN_ROOT}/skills/new/reference/local.md` (the commands),
`conventions.md` (what a good site looks like, and the checklist) and
`authoring.md` (the TOML and Liquid archival actually accepts). Read
`authoring.md` before touching a schema or a template; the syntax is specific.
## 1. Look first
```bash
bash "${CLAUDE_PLUGIN_ROOT}/bin/site-info.sh"
```
- `site=no`: this is not an Archival site. Hand over to the `archival:new`
skill rather than inventing one here.
- `site=legacy`: archival still reads `manifest.toml` and `objects.toml`. Do
not rename them unless asked.
- `archival=missing`: install it, never into the repo:
```bash
bash "${CLAUDE_PLUGIN_ROOT}/bin/install-archival.sh" "" "$HOME/.archival/bin"
```
then `export PATH="$HOME/.archival/bin:$PATH"` in each command that needs it.
- `remote=` and `logged_in=` decide how it deploys and whether media can be
uploaded (steps 5 and 6).
## 2. Baseline
```bash
export PATH="$HOME/.archival/bin:$PATH"
archival validate --json . && archival build .
```
A site that does not build before you touch it is reported to the person, not
silently repaired. Then read `archival_objects.toml`, `objects/`, `pages/`,
`layout/` and `archival_editor.toml` so the change fits what is there.
## 3. Make the change the Archival way
- **Content goes in objects.** A new paragraph, price, name or date is a field
on an object, never text in a `.liquid` file (conventions §1).
- **A new kind of thing** is: a type in `archival_objects.toml` → its files
under `objects/<type>/` → `pages/<type>.liquid` with `template = "<type>"` if
visitors click through to one → a `[[<type>.views]]` entry in
`archival_editor.toml`, and a shortcut if they will add to it often → a line
in `llms.txt` and `sitemap.xml` → `archival build`.
- **Changing a field's type**: the definition, then every object file, then the
templates that read it. `archival validate --json .` catches the objects.
- **Design** lives in `layout/`, `pages/` and `public/`.
- Never edit `dist/`, `.github/workflows/archival.yml`, or the manifest's
`upload_prefix`, `uploads_url`, `site_url` and `metadata`. The platform owns
those.
- Build after each file. The error names the file you just wrote.
## 4. Verify
```bash
rm -rf dist && archival build . && archival format .
```
then the checklist in conventions §11. `archival run . -p 1024` serves the site
at `http://localhost:1024` with live reload when the person wants to look. Say
that address is only on this computer.
## 5. Media
```bash
archival upload <type>/<filename> <field> <file> # a list item
archival upload <root> <field> <file> # a root object
```
It hashes and uploads the file and writes the `sha`/`filename`/`mime`/
`display_type` block into the object itself. It needs `logged_in=yes` (ask them
to run `archival login`, or run it in the background and relay the URL it
prints) and a remote it can read (`remote=archival` or `remote=github`), or
`-r domain/<domain>`. Never `git add` an image, a video or a download; a push
over 100 MB fails outright.
## 6. Commit and deploy
- `remote=archival`: `git push origin main` deploys.
- `remote=github` with `workflow=yes`: `git push` deploys through the site's
Actions workflow.
- `remote=none`: the folder is not connected to a hosted site yet. Follow
`local.md` G to list the person's sites and mint a remote, or F if the site
has never been published anywhere.
Deploy progress: `GET https://api.archival.dev/sites/github/<owner>/<name>`
with the login token (`local.md` F.6 shows how to read it without printing it),
under `setupInfo.deploy`.
Commit messages describe the change to the site, in the person's terms.
Wersje
| Repozytorium | Wynik | Licencja |
|---|---|---|
| Ta strona · archival-dev/archival | 100 | AGPL-3.0 |
| tammai/rspec | 90 | MIT |
Migawka bazy z 13.11.2026 · wersja paczki 3038f35e09f8 (skrót SHA-256 całej paczki)
Pobrania w ostatnich 30 dniach: 0 · ranking
Zgłoś problem z tym skillem: szkodliwa treść, błędny opis albo prośba autora o usunięcie.
Podobne skille
- 100new
Buduje stronę internetową w Archival na podstawie krótkiej rozmowy, odpowiada na pytania o działanie Archival i publikuje ją pod linkiem podglądu. - 100static-site-publish
Publikuje nowe strony statycznych serwisów przez commit do repozytorium GitHub, w jednym dozwolonym folderze treści. - 100launchdarkly-automations-archival-settings
Czyta i zmienia ustawienia automatycznego archiwizowania flag w LaunchDarkly Automations, np. okres karencji i zatwierdzenia. - 90wp-site
Prowadzi przez budowę strony WordPress od pustego folderu do działającej witryny i wskazuje, który etap jest następny. - 100hugo
Pomaga tworzyć strony i blogi w generatorze statycznych stron Hugo: motywy, wpisy z front matter, podgląd hugo server, konfiguracja hugo.toml i wdrożenie na GitHub Pages. - 100site-config-review
Przegląda konfigurację stencila.toml pod kątem poprawności, kompletności, przekierowań i gotowości do wdrożenia strony.