[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6PCG8Q6J394929AYXM9MB78":3,"skill-same:01M6PCG8Q6J394929AYXM9MB78":71},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":38,"linkOnly":39,"markdownContent":40,"markdownBody":41,"frontMatter":42,"refName":30,"commitSha":52,"skillDirectory":53,"forksCount":13,"repositoryHtmlUrl":54,"repositoryHomepageUrl":6,"ownerHtmlUrl":55,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":56,"install":57,"resources":70},"01M6PCG8Q6J394929AYXM9MB78","rag-infrastructure",null,"Build and operate Retrieval-Augmented Generation (RAG) infrastructure with vector stores, embedding pipelines, and hybrid search.","duc-duy-1612\u002FOnline-Delivery-Systems-Web","duc-duy-1612","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F192231384?v=4","01M6NSDKQQDKVTHZ8DZMZ1VYY6","01M6NSDKW8DDVTQHBT1HBDDJPT",0,"MIT","https:\u002F\u002Fgithub.com\u002Fduc-duy-1612\u002FOnline-Delivery-Systems-Web\u002Fblob\u002Fmain\u002F.agents\u002Fskills_archive\u002Frag-infrastructure\u002FSKILL.md",[],[18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37],"rag","infrastructure","rag infrastructure","hybrid","security","operate","relevant","Hybrid Search","embedding","scripts","critical","os","main","community","only","vector stores","access","devops-security","vector","search","2026-11-10 10:12:30",false,"---\nname: rag-infrastructure\ndescription: Build and operate Retrieval-Augmented Generation (RAG) infrastructure\n  with vector stores, embedding pipelines, and hybrid search.\ncategory: devops\nrisk: critical\nsource: https:\u002F\u002Fgithub.com\u002FBagelHole\u002FDevOps-Security-Agent-Skills\nsource_repo: BagelHole\u002FDevOps-Security-Agent-Skills\nsource_type: community\ndate_added: '2026-09-20'\nlicense: MIT\nlicense_source: https:\u002F\u002Fgithub.com\u002FBagelHole\u002FDevOps-Security-Agent-Skills\u002Fblob\u002Fmain\u002FLICENSE\ncompatibility: Requires the relevant OS\u002Fplatform tooling and privileged access where\n  noted. Docs-only; helper scripts and templates not bundled.\nmetadata:\n  author: devops-skills\n  version: '1.0'\n---\n\n# RAG Infrastructure\n\nProduction infrastructure for Retrieval-Augmented Generation: ingest documents, generate embeddings, store in vector databases, and serve grounded LLM responses.\n\n## When to Use This Skill\n\nUse this skill when:\n- Building a knowledge base Q&A system over internal documents\n- Implementing semantic search over large document collections\n- Reducing LLM hallucinations with retrieved context\n- Setting up embedding pipelines and vector store infrastructure\n- Deploying hybrid search (dense + sparse\u002FBM25)\n\n## Prerequisites\n\n- Python 3.10+ with `pip`\n- A vector database (Qdrant, Weaviate, Pinecone, or pgvector)\n- An embedding model (OpenAI, Cohere, or local via `sentence-transformers`)\n- An LLM endpoint (OpenAI API or self-hosted vLLM)\n- Docker for local vector DB deployment\n\n## Architecture Overview\n\n```\nDocuments → Chunker → Embedder → Vector Store\n                                      ↓\nUser Query → Embedder → Vector Store (search) → Reranker → LLM → Answer\n```\n\n## Embedding Pipeline\n\n```python\nfrom sentence_transformers import SentenceTransformer\nfrom qdrant_client import QdrantClient\nfrom qdrant_client.models import Distance, VectorParams, PointStruct\nimport uuid\n\n# Local embedding model (no API cost)\nmodel = SentenceTransformer(\"BAAI\u002Fbge-large-en-v1.5\")\n\n# Connect to Qdrant\nclient = QdrantClient(\"http:\u002F\u002Flocalhost:6333\")\n\n# Create collection\nclient.create_collection(\n    collection_name=\"knowledge-base\",\n    vectors_config=VectorParams(size=1024, distance=Distance.COSINE),\n)\n\ndef ingest_documents(docs: list[dict]):\n    \"\"\"Chunk, embed, and upsert documents.\"\"\"\n    points = []\n    for doc in docs:\n        chunks = chunk_text(doc[\"text\"], chunk_size=512, overlap=50)\n        embeddings = model.encode(chunks, batch_size=32, show_progress_bar=True)\n        for chunk, embedding in zip(chunks, embeddings):\n            points.append(PointStruct(\n                id=str(uuid.uuid4()),\n                vector=embedding.tolist(),\n                payload={\"text\": chunk, \"source\": doc[\"source\"], \"title\": doc[\"title\"]},\n            ))\n    client.upsert(collection_name=\"knowledge-base\", points=points)\n    print(f\"Ingested {len(points)} chunks\")\n```\n\n## Chunking Strategies\n\n```python\nfrom langchain.text_splitter import RecursiveCharacterTextSplitter\n\ndef chunk_text(text: str, chunk_size: int = 512, overlap: int = 50) -> list[str]:\n    \"\"\"Recursive character splitter — best general-purpose strategy.\"\"\"\n    splitter = RecursiveCharacterTextSplitter(\n        chunk_size=chunk_size,\n        chunk_overlap=overlap,\n        separators=[\"\\n\\n\", \"\\n\", \". \", \" \", \"\"],\n    )\n    return splitter.split_text(text)\n\n# For code\u002Fmarkdown — use language-aware splitter\nfrom langchain.text_splitter import MarkdownHeaderTextSplitter\n\nheaders = [(\"#\", \"H1\"), (\"##\", \"H2\"), (\"###\", \"H3\")]\nmd_splitter = MarkdownHeaderTextSplitter(headers_to_split_on=headers)\n```\n\n## Hybrid Search (Dense + Sparse)\n\n```python\nfrom qdrant_client.models import SparseVector, SparseVectorParams, NamedSparseVector\nfrom fastembed import SparseTextEmbedding\n\n# Qdrant hybrid collection (dense + BM25 sparse)\nclient.create_collection(\n    collection_name=\"hybrid-kb\",\n    vectors_config={\"dense\": VectorParams(size=1024, distance=Distance.COSINE)},\n    sparse_vectors_config={\"sparse\": SparseVectorParams()},\n)\n\nsparse_model = SparseTextEmbedding(\"prithivida\u002FSplade_PP_en_v1\")\n\ndef hybrid_search(query: str, top_k: int = 10) -> list[dict]:\n    dense_vec = model.encode(query).tolist()\n    sparse_vec = list(sparse_model.embed(query))[0]\n\n    results = client.query_points(\n        collection_name=\"hybrid-kb\",\n        prefetch=[\n            {\"query\": dense_vec, \"using\": \"dense\", \"limit\": 20},\n            {\"query\": SparseVector(indices=sparse_vec.indices.tolist(),\n                                   values=sparse_vec.values.tolist()),\n             \"using\": \"sparse\", \"limit\": 20},\n        ],\n        query={\"fusion\": \"rrf\"},   # Reciprocal Rank Fusion\n        limit=top_k,\n    )\n    return [{\"text\": p.payload[\"text\"], \"score\": p.score} for p in results.points]\n```\n\n## Reranking\n\n```python\nimport cohere\n\nco = cohere.Client(\"your-api-key\")\n\ndef rerank(query: str, candidates: list[str], top_n: int = 5) -> list[str]:\n    \"\"\"Rerank retrieved chunks for relevance (improves RAG quality ~20-30%).\"\"\"\n    response = co.rerank(\n        model=\"rerank-english-v3.0\",\n        query=query,\n        documents=candidates,\n        top_n=top_n,\n    )\n    return [candidates[r.index] for r in response.results]\n\n# Alternative: local reranker (no API cost)\nfrom sentence_transformers import CrossEncoder\nreranker = CrossEncoder(\"cross-encoder\u002Fms-marco-MiniLM-L-6-v2\")\n\ndef local_rerank(query: str, candidates: list[str], top_n: int = 5) -> list[str]:\n    pairs = [[query, c] for c in candidates]\n    scores = reranker.predict(pairs)\n    ranked = sorted(zip(candidates, scores), key=lambda x: x[1], reverse=True)\n    return [text for text, _ in ranked[:top_n]]\n```\n\n## RAG Query Pipeline\n\n```python\nfrom openai import OpenAI\n\nllm = OpenAI(base_url=\"http:\u002F\u002Flocalhost:8000\u002Fv1\", api_key=\"...\" # no real key needed for local endpoints)\n\ndef rag_query(user_question: str) -> str:\n    # 1. Retrieve\n    candidates = hybrid_search(user_question, top_k=20)\n    texts = [c[\"text\"] for c in candidates]\n\n    # 2. Rerank\n    top_chunks = local_rerank(user_question, texts, top_n=5)\n\n    # 3. Generate\n    context = \"\\n\\n---\\n\\n\".join(top_chunks)\n    response = llm.chat.completions.create(\n        model=\"meta-llama\u002FLlama-3.1-8B-Instruct\",\n        messages=[\n            {\"role\": \"system\", \"content\": (\n                \"Answer the question using only the provided context. \"\n                \"If the answer isn't in the context, say so.\\n\\nContext:\\n\" + context\n            )},\n            {\"role\": \"user\", \"content\": user_question},\n        ],\n        temperature=0.1,\n        max_tokens=1024,\n    )\n    return response.choices[0].message.content\n```\n\n## Docker Compose: Full RAG Stack\n\n```yaml\nservices:\n  qdrant:\n    image: qdrant\u002Fqdrant:latest\n    volumes:\n      - qdrant-data:\u002Fqdrant\u002Fstorage\n    ports:\n      - \"6333:6333\"\n    restart: unless-stopped\n\n  redis:\n    image: redis:7-alpine\n    volumes:\n      - redis-data:\u002Fdata\n    restart: unless-stopped\n\n  ingestion-worker:\n    build: .\u002Fingestion\n    environment:\n      - QDRANT_URL=http:\u002F\u002Fqdrant:6333\n      - REDIS_URL=redis:\u002F\u002Fredis:6379\n    depends_on: [qdrant, redis]\n    restart: unless-stopped\n\n  rag-api:\n    build: .\u002Fapi\n    ports:\n      - \"8080:8080\"\n    environment:\n      - QDRANT_URL=http:\u002F\u002Fqdrant:6333\n      - LLM_BASE_URL=http:\u002F\u002Fvllm:8000\u002Fv1\n    depends_on: [qdrant]\n    restart: unless-stopped\n\nvolumes:\n  qdrant-data:\n  redis-data:\n```\n\n## Common Issues\n\n| Issue | Cause | Fix |\n|-------|-------|-----|\n| Poor retrieval quality | Chunk size too large | Try 256–512 tokens; overlap 10–15% |\n| LLM ignores retrieved context | Context too long | Rerank and keep top 3–5 chunks |\n| Slow ingestion | Sequential embedding | Use `batch_size=64` and async upserts |\n| Stale documents | No re-ingestion pipeline | Track `doc_hash`; re-embed on change |\n| High embedding costs | All chunks re-embedded | Cache embeddings with hash-based dedup |\n\n## Best Practices\n\n- Use `BAAI\u002Fbge-large-en-v1.5` or `nomic-embed-text` for strong free embeddings.\n- Always rerank before passing to LLM — 5 precise chunks beat 20 noisy ones.\n- Store source metadata (URL, page, section) in vector payloads for citations.\n- Use namespace\u002Ftenant isolation in the vector store for multi-tenant RAG.\n- Evaluate with RAGAS metrics: faithfulness, answer relevancy, context precision.\n\n## Related Skills\n\n- vector-database-ops (`vector-database-ops`) - Qdrant\u002FWeaviate management\n- vllm-server (`vllm-server`) - Self-hosted LLM endpoint\n- ollama-stack (`ollama-stack`) - Local LLM for development\n- ai-pipeline-orchestration (`ai-pipeline-orchestration`) - Ingestion pipelines\n\n## Limitations\n\n- Infrastructure commands can disrupt services: confirm target host\u002Fscope and have backups\u002Fsnapshots before mutating state.\n- Docs-only import: upstream scripts and templates not bundled.\n\n","# RAG Infrastructure\n\nProduction infrastructure for Retrieval-Augmented Generation: ingest documents, generate embeddings, store in vector databases, and serve grounded LLM responses.\n\n## When to Use This Skill\n\nUse this skill when:\n- Building a knowledge base Q&A system over internal documents\n- Implementing semantic search over large document collections\n- Reducing LLM hallucinations with retrieved context\n- Setting up embedding pipelines and vector store infrastructure\n- Deploying hybrid search (dense + sparse\u002FBM25)\n\n## Prerequisites\n\n- Python 3.10+ with `pip`\n- A vector database (Qdrant, Weaviate, Pinecone, or pgvector)\n- An embedding model (OpenAI, Cohere, or local via `sentence-transformers`)\n- An LLM endpoint (OpenAI API or self-hosted vLLM)\n- Docker for local vector DB deployment\n\n## Architecture Overview\n\n```\nDocuments → Chunker → Embedder → Vector Store\n                                      ↓\nUser Query → Embedder → Vector Store (search) → Reranker → LLM → Answer\n```\n\n## Embedding Pipeline\n\n```python\nfrom sentence_transformers import SentenceTransformer\nfrom qdrant_client import QdrantClient\nfrom qdrant_client.models import Distance, VectorParams, PointStruct\nimport uuid\n\n# Local embedding model (no API cost)\nmodel = SentenceTransformer(\"BAAI\u002Fbge-large-en-v1.5\")\n\n# Connect to Qdrant\nclient = QdrantClient(\"http:\u002F\u002Flocalhost:6333\")\n\n# Create collection\nclient.create_collection(\n    collection_name=\"knowledge-base\",\n    vectors_config=VectorParams(size=1024, distance=Distance.COSINE),\n)\n\ndef ingest_documents(docs: list[dict]):\n    \"\"\"Chunk, embed, and upsert documents.\"\"\"\n    points = []\n    for doc in docs:\n        chunks = chunk_text(doc[\"text\"], chunk_size=512, overlap=50)\n        embeddings = model.encode(chunks, batch_size=32, show_progress_bar=True)\n        for chunk, embedding in zip(chunks, embeddings):\n            points.append(PointStruct(\n                id=str(uuid.uuid4()),\n                vector=embedding.tolist(),\n                payload={\"text\": chunk, \"source\": doc[\"source\"], \"title\": doc[\"title\"]},\n            ))\n    client.upsert(collection_name=\"knowledge-base\", points=points)\n    print(f\"Ingested {len(points)} chunks\")\n```\n\n## Chunking Strategies\n\n```python\nfrom langchain.text_splitter import RecursiveCharacterTextSplitter\n\ndef chunk_text(text: str, chunk_size: int = 512, overlap: int = 50) -> list[str]:\n    \"\"\"Recursive character splitter — best general-purpose strategy.\"\"\"\n    splitter = RecursiveCharacterTextSplitter(\n        chunk_size=chunk_size,\n        chunk_overlap=overlap,\n        separators=[\"\\n\\n\", \"\\n\", \". \", \" \", \"\"],\n    )\n    return splitter.split_text(text)\n\n# For code\u002Fmarkdown — use language-aware splitter\nfrom langchain.text_splitter import MarkdownHeaderTextSplitter\n\nheaders = [(\"#\", \"H1\"), (\"##\", \"H2\"), (\"###\", \"H3\")]\nmd_splitter = MarkdownHeaderTextSplitter(headers_to_split_on=headers)\n```\n\n## Hybrid Search (Dense + Sparse)\n\n```python\nfrom qdrant_client.models import SparseVector, SparseVectorParams, NamedSparseVector\nfrom fastembed import SparseTextEmbedding\n\n# Qdrant hybrid collection (dense + BM25 sparse)\nclient.create_collection(\n    collection_name=\"hybrid-kb\",\n    vectors_config={\"dense\": VectorParams(size=1024, distance=Distance.COSINE)},\n    sparse_vectors_config={\"sparse\": SparseVectorParams()},\n)\n\nsparse_model = SparseTextEmbedding(\"prithivida\u002FSplade_PP_en_v1\")\n\ndef hybrid_search(query: str, top_k: int = 10) -> list[dict]:\n    dense_vec = model.encode(query).tolist()\n    sparse_vec = list(sparse_model.embed(query))[0]\n\n    results = client.query_points(\n        collection_name=\"hybrid-kb\",\n        prefetch=[\n            {\"query\": dense_vec, \"using\": \"dense\", \"limit\": 20},\n            {\"query\": SparseVector(indices=sparse_vec.indices.tolist(),\n                                   values=sparse_vec.values.tolist()),\n             \"using\": \"sparse\", \"limit\": 20},\n        ],\n        query={\"fusion\": \"rrf\"},   # Reciprocal Rank Fusion\n        limit=top_k,\n    )\n    return [{\"text\": p.payload[\"text\"], \"score\": p.score} for p in results.points]\n```\n\n## Reranking\n\n```python\nimport cohere\n\nco = cohere.Client(\"your-api-key\")\n\ndef rerank(query: str, candidates: list[str], top_n: int = 5) -> list[str]:\n    \"\"\"Rerank retrieved chunks for relevance (improves RAG quality ~20-30%).\"\"\"\n    response = co.rerank(\n        model=\"rerank-english-v3.0\",\n        query=query,\n        documents=candidates,\n        top_n=top_n,\n    )\n    return [candidates[r.index] for r in response.results]\n\n# Alternative: local reranker (no API cost)\nfrom sentence_transformers import CrossEncoder\nreranker = CrossEncoder(\"cross-encoder\u002Fms-marco-MiniLM-L-6-v2\")\n\ndef local_rerank(query: str, candidates: list[str], top_n: int = 5) -> list[str]:\n    pairs = [[query, c] for c in candidates]\n    scores = reranker.predict(pairs)\n    ranked = sorted(zip(candidates, scores), key=lambda x: x[1], reverse=True)\n    return [text for text, _ in ranked[:top_n]]\n```\n\n## RAG Query Pipeline\n\n```python\nfrom openai import OpenAI\n\nllm = OpenAI(base_url=\"http:\u002F\u002Flocalhost:8000\u002Fv1\", api_key=\"...\" # no real key needed for local endpoints)\n\ndef rag_query(user_question: str) -> str:\n    # 1. Retrieve\n    candidates = hybrid_search(user_question, top_k=20)\n    texts = [c[\"text\"] for c in candidates]\n\n    # 2. Rerank\n    top_chunks = local_rerank(user_question, texts, top_n=5)\n\n    # 3. Generate\n    context = \"\\n\\n---\\n\\n\".join(top_chunks)\n    response = llm.chat.completions.create(\n        model=\"meta-llama\u002FLlama-3.1-8B-Instruct\",\n        messages=[\n            {\"role\": \"system\", \"content\": (\n                \"Answer the question using only the provided context. \"\n                \"If the answer isn't in the context, say so.\\n\\nContext:\\n\" + context\n            )},\n            {\"role\": \"user\", \"content\": user_question},\n        ],\n        temperature=0.1,\n        max_tokens=1024,\n    )\n    return response.choices[0].message.content\n```\n\n## Docker Compose: Full RAG Stack\n\n```yaml\nservices:\n  qdrant:\n    image: qdrant\u002Fqdrant:latest\n    volumes:\n      - qdrant-data:\u002Fqdrant\u002Fstorage\n    ports:\n      - \"6333:6333\"\n    restart: unless-stopped\n\n  redis:\n    image: redis:7-alpine\n    volumes:\n      - redis-data:\u002Fdata\n    restart: unless-stopped\n\n  ingestion-worker:\n    build: .\u002Fingestion\n    environment:\n      - QDRANT_URL=http:\u002F\u002Fqdrant:6333\n      - REDIS_URL=redis:\u002F\u002Fredis:6379\n    depends_on: [qdrant, redis]\n    restart: unless-stopped\n\n  rag-api:\n    build: .\u002Fapi\n    ports:\n      - \"8080:8080\"\n    environment:\n      - QDRANT_URL=http:\u002F\u002Fqdrant:6333\n      - LLM_BASE_URL=http:\u002F\u002Fvllm:8000\u002Fv1\n    depends_on: [qdrant]\n    restart: unless-stopped\n\nvolumes:\n  qdrant-data:\n  redis-data:\n```\n\n## Common Issues\n\n| Issue | Cause | Fix |\n|-------|-------|-----|\n| Poor retrieval quality | Chunk size too large | Try 256–512 tokens; overlap 10–15% |\n| LLM ignores retrieved context | Context too long | Rerank and keep top 3–5 chunks |\n| Slow ingestion | Sequential embedding | Use `batch_size=64` and async upserts |\n| Stale documents | No re-ingestion pipeline | Track `doc_hash`; re-embed on change |\n| High embedding costs | All chunks re-embedded | Cache embeddings with hash-based dedup |\n\n## Best Practices\n\n- Use `BAAI\u002Fbge-large-en-v1.5` or `nomic-embed-text` for strong free embeddings.\n- Always rerank before passing to LLM — 5 precise chunks beat 20 noisy ones.\n- Store source metadata (URL, page, section) in vector payloads for citations.\n- Use namespace\u002Ftenant isolation in the vector store for multi-tenant RAG.\n- Evaluate with RAGAS metrics: faithfulness, answer relevancy, context precision.\n\n## Related Skills\n\n- vector-database-ops (`vector-database-ops`) - Qdrant\u002FWeaviate management\n- vllm-server (`vllm-server`) - Self-hosted LLM endpoint\n- ollama-stack (`ollama-stack`) - Local LLM for development\n- ai-pipeline-orchestration (`ai-pipeline-orchestration`) - Ingestion pipelines\n\n## Limitations\n\n- Infrastructure commands can disrupt services: confirm target host\u002Fscope and have backups\u002Fsnapshots before mutating state.\n- Docs-only import: upstream scripts and templates not bundled.\n\n",{"name":5,"risk":28,"source":43,"license":14,"category":44,"metadata":45,"date_added":48,"description":7,"source_repo":49,"source_type":31,"compatibility":50,"license_source":51},"https:\u002F\u002Fgithub.com\u002FBagelHole\u002FDevOps-Security-Agent-Skills","devops",{"author":46,"version":47},"devops-skills","1.0","2026-09-20","BagelHole\u002FDevOps-Security-Agent-Skills","Requires the relevant OS\u002Fplatform tooling and privileged access where noted. Docs-only; helper scripts and templates not bundled.","https:\u002F\u002Fgithub.com\u002FBagelHole\u002FDevOps-Security-Agent-Skills\u002Fblob\u002Fmain\u002FLICENSE","f3312ba749561b6159f477dbfc9313c875d40a56",".agents\u002Fskills_archive\u002Frag-infrastructure","https:\u002F\u002Fgithub.com\u002Fduc-duy-1612\u002FOnline-Delivery-Systems-Web","https:\u002F\u002Fgithub.com\u002Fduc-duy-1612","2026-10-25 23:50:58",{"source":8,"skillName":5,"commands":58,"prompt":68,"zipAvailable":69},[59,62,65],{"tool":60,"command":61},"npx","npx skills add duc-duy-1612\u002FOnline-Delivery-Systems-Web --skill rag-infrastructure",{"tool":63,"command":64},"bunx","bunx skills add duc-duy-1612\u002FOnline-Delivery-Systems-Web --skill rag-infrastructure",{"tool":66,"command":67},"pnpm","pnpm dlx skills add duc-duy-1612\u002FOnline-Delivery-Systems-Web --skill rag-infrastructure","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fduc-duy-1612\u002FOnline-Delivery-Systems-Web\u002Fblob\u002Fmain\u002F.agents\u002Fskills_archive\u002Frag-infrastructure\u002FSKILL.md",true,[],[72,95,119,144,171,195],{"skillNo":73,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":76,"categories":77,"keywords":83,"publishedAt":38},"01M6Q0FA5FZ410Z33Z3TFQ4DBA","zero-trust","Implement zero-trust network architecture. Configure identity-based access, micro-segmentation, and continuous verification. Use when implementing modern security architectures.","https:\u002F\u002Fgithub.com\u002Fduc-duy-1612\u002FOnline-Delivery-Systems-Web\u002Fblob\u002Fmain\u002F.agents\u002Fskills_archive\u002Fzero-trust\u002FSKILL.md",[78],{"categoryNo":79,"categoryName":80,"taxonomyType":81,"taxonomyTypeDesc":82},"01JBSEEDCAT000000000000011","运维与部署","SCENARIO","使用场景",[84,85,86,87,88,89,74,90,35,34,91,32,31,30,92,28,93,27,24,94],"ZeRO","trust","zero trust","architecture","assessment","verification","segmentation","configure","active","scope","implementing",{"skillNo":96,"name":97,"nameZh":6,"description":98,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":99,"categories":100,"keywords":101,"publishedAt":38},"01M6Q0F1YZJW3KX5FBKEVZ8CV4","zcode-delegate","Delegate coding tasks to the Z.AI ZCode CLI only when the user explicitly requests it, while the orchestrator retains review and landing responsibility.","https:\u002F\u002Fgithub.com\u002Fduc-duy-1612\u002FOnline-Delivery-Systems-Web\u002Fblob\u002Fmain\u002F.agents\u002Fskills_archive\u002Fzcode-delegate\u002FSKILL.md",[],[102,103,104,105,106,107,108,109,110,111,112,113,114,32,115,31,116,117,28,118],"delegate","zcode","zcode delegate","review","cli","orchestration","tasks","run","Bash","model","node","read","orchestrator","files","desktop app","app","agent-orchestration",{"skillNo":120,"name":121,"nameZh":6,"description":122,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":123,"categories":124,"keywords":125,"publishedAt":38},"01M6Q0ETK41VGNMYEXW0T00AJN","youtube-transcript","Fetch YouTube transcripts through DeepAPI or local fallback tooling and save clean text output.","https:\u002F\u002Fgithub.com\u002Fduc-duy-1612\u002FOnline-Delivery-Systems-Web\u002Fblob\u002Fmain\u002F.agents\u002Fskills_archive\u002Fyoutube-transcript\u002FSKILL.md",[],[126,127,128,129,130,131,132,133,134,135,136,137,138,31,30,139,140,141,142,143],"transcript","youtube transcript","youtube","safe","github","codex","research","license","https","text","fetch","tooling","clean","local","output","deepapi","youtube transcripts","blob",{"skillNo":145,"name":146,"nameZh":6,"description":147,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":148,"categories":149,"keywords":161,"publishedAt":38},"01M6Q0EPPH2PVMDW40017F8WX0","youtube-transcript-skills","Fetch YouTube video transcripts, search videos\u002Fchannels, browse channels, and extract playlists via the getyoutubetranscript.com API - free tier, no card required.","https:\u002F\u002Fgithub.com\u002Fduc-duy-1612\u002FOnline-Delivery-Systems-Web\u002Fblob\u002Fmain\u002F.agents\u002Fskills_archive\u002Fyoutube-transcript-skills\u002FSKILL.md",[150,153,156],{"categoryNo":151,"categoryName":152,"taxonomyType":81,"taxonomyTypeDesc":82},"01JBSEEDCAT000000000000010","后端与接口",{"categoryNo":154,"categoryName":155,"taxonomyType":81,"taxonomyTypeDesc":82},"01JBSEEDCAT000000000000016","图像与多媒体",{"categoryNo":157,"categoryName":158,"taxonomyType":159,"taxonomyTypeDesc":160},"01JBSEEDCAT000000000000030","开发工程师","PROFESSION","适用职业",[126,127,128,162,129,163,130,131,133,37,164,165,166,136,167,31,30,168,169,170],"youtube transcript skills","gemini","api","Video","browse","api-integration","free","card","channels",{"skillNo":172,"name":173,"nameZh":6,"description":174,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":175,"categories":176,"keywords":181,"publishedAt":38},"01M6Q0EAB96PX83W5PTE8A0R8D","youtube-notetaker","Turn YouTube talks into local study notes with slides, transcripts, editable annotations, and a markdown-backed viewer.","https:\u002F\u002Fgithub.com\u002Fduc-duy-1612\u002FOnline-Delivery-Systems-Web\u002Fblob\u002Fmain\u002F.agents\u002Fskills_archive\u002Fyoutube-notetaker\u002FSKILL.md",[177,180],{"categoryNo":178,"categoryName":179,"taxonomyType":81,"taxonomyTypeDesc":82},"01JBSEEDCAT000000000000003","演示文稿",{"categoryNo":154,"categoryName":155,"taxonomyType":81,"taxonomyTypeDesc":82},[128,182,183,106,129,184,185,186,187,133,165,188,189,190,191,30,192,139,193,194],"youtube notetaker","notetaker","workflow","official","notes","code","markdown","slides","annotations","README","study","dair-academy","viewer",{"skillNo":196,"name":197,"nameZh":6,"description":198,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":199,"categories":200,"keywords":202,"publishedAt":38},"01M6Q0E55F0GW5MX48V8HXS8B5","youtube-full","Fetch YouTube transcripts, search videos, browse channels, and extract playlists via TranscriptAPI — no yt-dlp, no Google API key, works from any cloud server.","https:\u002F\u002Fgithub.com\u002Fduc-duy-1612\u002FOnline-Delivery-Systems-Web\u002Fblob\u002Fmain\u002F.agents\u002Fskills_archive\u002Fyoutube-full\u002FSKILL.md",[201],{"categoryNo":151,"categoryName":152,"taxonomyType":81,"taxonomyTypeDesc":82},[203,128,204,205,129,163,37,206,207,164,165,166,136,208,167,31,30,168,209,142],"youtube full","full","google","server","dlp","credits","automatic"]