VSS (Vibestream Syndication) is a protocol for syndicating software the way RSS syndicates writing: a vibestream is a plain-text feed of vibeware items that readers can track, install, fork, or recommend without needing a central registry.
VSS
VSS (Vibestream Syndication), not yet assigned a GSV registry designation, is a protocol design (infrastructure / media-ecology pace-layer) for syndicating small software objects the way RSS syndicates writing. Its core unit is the vibestream, a plain-text feed of vibeware items — micro-apps, prompts, Claude skills, MCP servers, workflows, templates, scripts, native apps, firmware, agent toolkits — that a reader can list, filter, install, fork, recommend, or watch. The protocol's central bet is architectural: as small software-adjacent objects become as cheap to publish as blog posts, the missing layer isn't another centralized marketplace but a decentralized interop format, where a publisher hosts a plain-text feed at their own URL and readers, launchers, crawlers, search engines, and recommenders grow around it independently — deliberately reproducing the mechanism that let RSS spread rather than a walled-garden app store. The key design decision is keeping the adoption floor minimal: a valid feed needs only a version tag, title, and feed URL, with items carrying `id`, `kind`, `title`, and `url` — everything else stays optional until a reader actually needs it, which keeps a feed hand-writable by design. The seed repository backs this with a v0.1 spec sketch, a provisional JSON Schema, and two dependency-free Python tools — a validator enforcing the minimal floor while tolerating unknown extension fields, and a reader prototype that lists and filters validated items by kind. As a protocol-fiction piece it sits in the infrastructure/media-ecology layer of the GSV portfolio alongside the broader vibeware ecosystem it presupposes (skills, MCP servers, agent toolkits) but is not documented as linked to a specific sibling project. Status: this is a private, early design-doc-stage seed — a spec sketch, schema, and two small tools exist, but no live feeds, registries, or reader/crawler ecosystem have been built.