Update to Vibe Coding

{
  "description": "Audit, fix or verify cache keys, invalidation, TTL, stale reads and coherency across authoritative writes and readers. Focus on cache correctness rather than Redis queue semantics.",
  "included_files": [
    {
      "relative_path": "agents/openai.yaml",
      "size_in_bytes": 212
    },
    {
      "relative_path": "references/cache-coherency-check.md",
      "size_in_bytes": 995
    },
    {
      "relative_path": "references/cache-invalidation-audit.md",
      "size_in_bytes": 1185
    },
    {
      "relative_path": "references/cache-invalidation-polish.md",
      "size_in_bytes": 1694
    }
  ],
  "name": "vibe-cache",
  "skill_md_contents": "---\nname: vibe-cache\ndescription: \"Audit, fix or verify cache keys, invalidation, TTL, stale reads and coherency across authoritative writes and readers. Focus on cache correctness rather than Redis queue semantics.\"\n---\n\n# Caching & freshness\n\nRead [the shared workflow](../../references/workflow.md) once per task, then the one recipe matching the requested operation and boundary. Do not load every recipe. The user's request controls scope and whether edits are allowed.\n\n## Recipes\n\n| Recipe | Operation |\n|---|---|\n| [Cache Coherency Check](references/cache-coherency-check.md) | check |\n| [Cache Invalidation Audit](references/cache-invalidation-audit.md) | audit |\n| [Cache Invalidation Polish](references/cache-invalidation-polish.md) | implement |\n\nFor an explicit multi-stage request, follow the necessary stages without discarding requested work. For a single stage, deliver that result and stop. Use [the catalog](../../references/catalog.md) only if the requested boundary belongs elsewhere.\n\nClient optimistic state and reconnect belong to [Realtime](../vibe-realtime/SKILL.md); server rendering and serialized responses belong to [Web](../vibe-web/SKILL.md). This skill owns shared cache identity, invalidation and freshness.\n"
}