Update to Designly

{
  "name": "reference-memory",
  "description": "Local-first reference memory and scoped preference ledger manager. This skill should be used when saving, recalling, updating, or deleting reference records with stable REF IDs (e.g. REF-1042), managing user likes/dislikes, or querying persistent taste preferences.",
  "included_files": [
    {
      "relative_path": "agents/openai.yaml",
      "size_in_bytes": 375
    },
    {
      "relative_path": "assets/large-logo.svg",
      "size_in_bytes": 739
    },
    {
      "relative_path": "assets/reference-memory.template.json",
      "size_in_bytes": 82
    },
    {
      "relative_path": "assets/small-logo.svg",
      "size_in_bytes": 739
    },
    {
      "relative_path": "schemas/reference-memory.schema.json",
      "size_in_bytes": 2353
    },
    {
      "relative_path": "scripts/reference_memory.py",
      "size_in_bytes": 6647
    },
    {
      "relative_path": "scripts/test_taste_memory.py",
      "size_in_bytes": 1796
    }
  ],
  "skill_md_contents": "---\nname: reference-memory\ndescription: Local-first reference memory and scoped preference ledger manager. This skill should be used when saving, recalling, updating, or deleting reference records with stable REF IDs (e.g. REF-1042), managing user likes/dislikes, or querying persistent taste preferences.\n---\n\n# Reference Memory\n\nReference Memory provides local-first, transparent persistence for user references, assigned design jobs, and scoped feedback ledgers. It uses stable human-readable IDs (`REF-####`) and does not make misleading claims of fine-tuning or modifying underlying model weights.\n\n---\n\n## 1. Core Workflow\n\n1. **Reference Registration**:\n   - Assign next available sequential ID: `REF-####` (e.g. `REF-1001`, `REF-1002`).\n   - Store reference metadata, file path, source tags, and primary design jobs (e.g. `lighting`, `composition`, `palette`).\n\n2. **Scoped Preference Feedback**:\n   - When a user expresses a preference (e.g., \"I love the rim light in REF-1042 but hate the typography\"), scope the feedback precisely:\n     - Record `likes: [\"lighting.rim_light\"]`\n     - Record `dislikes: [\"typography.layout\"]`\n   - Never allow a dislike on one dimension (typography) to discard or invalidate rules on another dimension (lighting).\n\n3. **Recall & Filtering**:\n   - Query references by job tag, domain (e.g. `fmcg`, `luxury`, `automotive`), or aesthetic keyword.\n   - Return active Taste Profiles associated with the recalled reference.\n\n4. **Output Contract**:\n   - Emits structured JSON reference records and `DesignSignalPacket` containing `taste_state.reference_memory`.\n\n---\n\n## 2. CLI Tooling\n\nExecute the local Reference Memory CLI:\n```bash\n# Add a new reference\npython3 scripts/reference_memory.py add --id REF-1001 --title \"Nordic Minimalist Poster\" --jobs composition,palette\n\n# Query existing references\npython3 scripts/reference_memory.py list\n\n# Recall specific reference\npython3 scripts/reference_memory.py get REF-1001\n\n# Record scoped feedback\npython3 scripts/reference_memory.py feedback REF-1001 --like \"lighting.soft_directional\" --dislike \"background.clutter\"\n```\n\n---\n\n## 3. Cross-Skill Neural Connections & References\n\n### Peer & Downstream Skills\n- [Taste Engine](../taste-engine/SKILL.md) — Extracting transferable rules from recalled references\n- [Brand Intelligence](../brand-intelligence/SKILL.md) — Auditing reference tags against brand constraints\n- [Designly Director](../designly-director/SKILL.md) — Lead orchestrator and memory state manager\n\n### Schemas & References\n- [Reference Memory Schema](schemas/reference-memory.schema.json) — Local schema\n- [Reference Memory Guide](../../shared/references/reference-memory.md) — Usage rules\n- [Signal Packet Schema](../../shared/contracts/signal-packet.schema.json) — Neural Mesh handoff\n"
}