# Listing Rulebook

> Check a B2B app marketplace's listing-content rulebook against objective, self-checkable standards, and lint a
> sample listing against it. Paste the draft rules and one listing; the browser reads every rule as objective or
> principles-only, checks the five rulebook gates, ranks the strictness, media, taxonomy, localization, badge and decay
> rungs from a short interview and lints the listing for the named rejection triggers. Metered runs write the
> published rulebook, and a review of the listing. Built on the SkillSafe skill @samber/app-marketplace-listing-standards,
> from samber/developer-platform-skills (MIT), by Samuel Berthe.

https://listing-rulebook.skillsafe.ai/

## What is free
- The rulebook gates: an objective reject criterion on every required field (name, hook, body, screenshots, categories, landing page, support, privacy), exactly one geometry per visual asset, a hard per-listing category cap, decay triggers with an owner and a consequence, a display location and currency rule for every badge; a naming-order rule, a video ceiling set for who evaluates, and no external certification claimed.
- The rungs: field-spec strictness, media, taxonomy, localization, badge display and decay enforcement, ranked from who maintains and who evaluates listings, volume, releases, languages, badges, deadline, mandate, effort and the assets already owned.
- The listing lint: the draft's own length limits, the platform's name before the app's, any term repeated past the count, quoted testimonials with no named source, guarantee claims, a missing, PDF or repository landing page, support and privacy, screenshot geometry and count, the video ceiling and the category cap.
- A submitter self-check, a note to the listing's vendor naming each failed trigger and its fix, and exports: the listing lint (.csv), the checks (.csv), everything (.json).

## The metered lanes
Two tasks on gpt-terra, metered per run, routed by `task`. Input string fields: `task` (rulebook or review), `market`, `draft`, `carry` (review only: the rulebook's decisions) and `facts`.
- `rulebook` replies with `summary`, `rungs`, `fields` (each with a requirement and a self-checkable reject criterion), `media`, `triggers` (all nine), `taxonomy`, `decay`, `badges`, `runway` and `caveats`.
- `review` replies with `summary`, `verdict`, `violations` (each with a trigger id or judgment, the field, the finding and the fix), `passes` and `caveats`.
The page holds every reply to its own facts: the ranked rungs, every core field with a self-checkable criterion, one geometry per asset, all nine triggers, the cap, complete staleness and badge rules, a runway and no external certification; a verdict that follows the page, every page violation cited once by trigger and field, never a trigger the page passed.

## Links
- API tutorial: https://listing-rulebook.skillsafe.ai/api.html
- Source skill: https://skillsafe.ai/skill/@samber/app-marketplace-listing-standards