{"id":67715,"plugin_id":"plugin_asdk_app_6aca33a416588191b9ee30c231d41940","kind":"skill","collection_source":"plugin_package","comparison_source":null,"observed_at":"2026-10-28T13:10:43.573Z","digest":"2a0b83211482c9d67bc25e985c642c3ca4c4b574ed81eec81933d193c0df30c6","against":null,"payload":{"description":"Inspect the live Givon AI catalog, model schemas, parameters, capabilities and costs before generation. Task-specific selection guidance lives in model-selection-guidance.","included_files":[{"relative_path":"agents/openai.yaml","size_in_bytes":380},{"relative_path":"references/model-selection.md","size_in_bytes":1880}],"name":"givon-ai-model-router","skill_md_contents":"---\nname: givon-ai-model-router\ndescription: \"Inspect the live Givon AI catalog, model schemas, parameters, capabilities and costs before generation. Task-specific selection guidance lives in model-selection-guidance.\"\nmetadata:\n  version: skill-render-ea6bd93365667f6d\n---\n\n# Givon AI Model Router\n\nRecommend one practical model path by default. Keep catalog detail hidden unless\nthe user asks for comparison.\n\n## Route\n\n1. Start with `givon_list_models`; filter by modality, status, tags, or query.\n2. Short-list active models only unless the user asks about coming-soon or\n   deprecated models.\n3. For every candidate you might recommend, call `givon_get_model_schema`.\n4. Compare the user's required fields with schema-supported fields, asset slots,\n   variants, and cost metadata.\n5. Recommend one model and the minimum valid `input` shape.\n\nBefore live catalog/schema inspection, name only the requested result and\ncompatibility needs. Do not declare an executable generation/edit mode:\nreference-conditioned creation may be exposed as a still-generation input or\nas an edit operation, and the selected schema owns that distinction.\n\n## CLI Flow\n\nWhen working through the CLI, use `givon model list --kind <image|video>` for\nlive catalog discovery and `givon model get <id>` for schema inspection. Choose\nthe model from these instructions and the live schema; do not rely on a separate\nCLI heuristic.\n\nSelection priorities come from the user's requested result and the relevant\nspecialist workflow. Availability, input compatibility, quality guidance,\ndefaults and cost come from the live catalog and schema.\n\n## Answer Shape\n\n- Give one recommendation first.\n- Use a table only for explicit comparisons.\n- Mention costs from model metadata or live responses.\n- A separate creation intent starts generation; catalog questions end with the\n  recommendation and supported input shape.\n- A user-named model gets a full-catalog lookup when filtered search misses it.\n- Classify by the requested result, required inputs and user priorities, not by\n  tags alone.\n- For asset/reference workflows, treat `x-assetSlots` as capability discovery\n  plus a schema checklist. Media is passed only through fields present in the\n  selected model's `input` schema; a schema without media fields is prompt-only.\n\nRead [references/model-selection.md](references/model-selection.md) for the\nschema checklist.\n"},"changes":[],"summary":"First saved snapshot. No earlier version is available for comparison.","summary_kind":"deterministic","summary_metadata":{}}