voyageai-cli 1.30.0 → 1.30.2

package/src/workflows/tests/enrich-and-ingest.happy-path.test.json

@@ -0,0 +1,38 @@
+{
+  "name": "enrich-and-ingest: happy path",
+  "inputs": {
+    "document_url": "https://api.example.com/documents/123/metadata",
+    "text": "This is the document text to ingest after enrichment with external metadata."
+  },
+  "mocks": {
+    "http": {
+      "status": 200,
+      "statusText": "OK",
+      "headers": { "content-type": "application/json" },
+      "body": {
+        "title": "Getting Started Guide",
+        "author": "Jane Developer",
+        "category": "tutorials",
+        "timestamp": "2025-01-15T10:00:00Z"
+      },
+      "durationMs": 150
+    },
+    "ingest": {
+      "chunks": 2,
+      "source": "Getting Started Guide",
+      "collection": "default"
+    }
+  },
+  "expect": {
+    "steps": {
+      "fetch_metadata": { "status": "completed" },
+      "store": { "status": "completed" },
+      "notify": { "status": "completed" }
+    },
+    "output": {
+      "source": { "type": "string", "minLength": 1 },
+      "chunks": { "type": "number" }
+    },
+    "noErrors": true
+  }
+}

package/src/workflows/tests/enrich-and-ingest.notify-fails.test.json

@@ -0,0 +1,38 @@
+{
+  "name": "enrich-and-ingest: notification failure (continueOnError)",
+  "inputs": {
+    "document_url": "https://api.example.com/documents/456/metadata",
+    "text": "Document text for ingestion even if notification fails."
+  },
+  "mocks": {
+    "http": {
+      "status": 200,
+      "statusText": "OK",
+      "headers": { "content-type": "application/json" },
+      "body": {
+        "title": "API Reference",
+        "author": "John Dev",
+        "category": "reference",
+        "timestamp": "2025-02-01T14:00:00Z"
+      },
+      "durationMs": 200
+    },
+    "ingest": {
+      "chunks": 5,
+      "source": "API Reference",
+      "collection": "default"
+    }
+  },
+  "expect": {
+    "steps": {
+      "fetch_metadata": { "status": "completed" },
+      "store": { "status": "completed" },
+      "notify": { "status": "completed" }
+    },
+    "output": {
+      "source": { "type": "string", "minLength": 1 },
+      "chunks": { "type": "number" }
+    },
+    "noErrors": true
+  }
+}

package/src/workflows/tests/intelligent-ingest.all-filtered.test.json

@@ -0,0 +1,26 @@
+{
+  "name": "intelligent-ingest: all chunks filtered out (too short)",
+  "inputs": {
+    "text": "Short. Tiny. Small text.",
+    "source": "tiny.md",
+    "similarity_threshold": 0.92
+  },
+  "mocks": {
+    "search": {
+      "results": [],
+      "resultCount": 0
+    }
+  },
+  "expect": {
+    "steps": {
+      "split": { "status": "completed" },
+      "filter_short": { "status": "completed" },
+      "check_each": { "status": "completed" },
+      "embed_novel": { "status": "completed" }
+    },
+    "output": {
+      "summary": { "type": "string" }
+    },
+    "noErrors": true
+  }
+}

package/src/workflows/tests/intelligent-ingest.happy-path.test.json

@@ -0,0 +1,28 @@
+{
+  "name": "intelligent-ingest: happy path",
+  "inputs": {
+    "text": "This is a long document about vector databases and how they work. Vector databases store high-dimensional embeddings and support approximate nearest neighbor search for fast retrieval. They are essential for modern RAG applications. The key advantage is that semantic similarity can be computed efficiently at scale. Various indexing strategies like HNSW and IVF are used to speed up queries.",
+    "source": "vector-db-guide.md",
+    "similarity_threshold": 0.92
+  },
+  "mocks": {
+    "search": {
+      "results": [
+        { "text": "Some existing doc about databases", "score": 0.45 }
+      ],
+      "resultCount": 1
+    }
+  },
+  "expect": {
+    "steps": {
+      "split": { "status": "completed" },
+      "filter_short": { "status": "completed" },
+      "check_each": { "status": "completed" },
+      "embed_novel": { "status": "completed" }
+    },
+    "output": {
+      "summary": { "type": "string", "minLength": 1 }
+    },
+    "noErrors": true
+  }
+}

package/src/workflows/tests/kb-health-report.custom-queries.test.json

@@ -0,0 +1,24 @@
+{
+  "name": "kb-health-report: custom queries",
+  "inputs": {
+    "test_queries": ["vector search", "embeddings"]
+  },
+  "mocks": {
+    "query": {
+      "results": [],
+      "resultCount": 0
+    },
+    "generate": {
+      "text": "Knowledge Base Health Report: 2 topics tested. Low coverage detected across test queries.",
+      "model": "claude-3-haiku",
+      "tokensUsed": 60
+    }
+  },
+  "expect": {
+    "steps": {
+      "coverage_check": { "status": "completed" },
+      "report": { "status": "completed" }
+    },
+    "noErrors": true
+  }
+}

package/src/workflows/tests/kb-health-report.happy-path.test.json

@@ -0,0 +1,26 @@
+{
+  "name": "kb-health-report: happy path",
+  "inputs": {
+    "test_queries": ["authentication", "deployment", "error handling"]
+  },
+  "mocks": {
+    "query": {
+      "results": [
+        { "text": "Auth docs available", "score": 0.9 }
+      ],
+      "resultCount": 1
+    },
+    "generate": {
+      "text": "Knowledge Base Health Report: 3 topics tested. Coverage looks good with results found for all test queries.",
+      "model": "claude-3-haiku",
+      "tokensUsed": 85
+    }
+  },
+  "expect": {
+    "steps": {
+      "coverage_check": { "status": "completed" },
+      "report": { "status": "completed" }
+    },
+    "noErrors": true
+  }
+}

package/src/workflows/tests/multi-collection-search.happy-path.test.json

@@ -0,0 +1,40 @@
+{
+  "name": "multi-collection-search: happy path",
+  "inputs": {
+    "query": "how to deploy",
+    "collection1": "api_docs",
+    "collection2": "tutorials",
+    "limit": 5
+  },
+  "mocks": {
+    "query": {
+      "results": [
+        { "text": "Deploy using Docker containers", "score": 0.92, "source": "deploy.md" },
+        { "text": "CI/CD pipeline setup guide", "score": 0.85, "source": "ci.md" }
+      ],
+      "resultCount": 2
+    },
+    "rerank": {
+      "results": [
+        { "text": "Deploy using Docker containers", "score": 0.96, "source": "deploy.md" },
+        { "text": "CI/CD pipeline setup guide", "score": 0.91, "source": "ci.md" },
+        { "text": "Deploy using Docker containers", "score": 0.89, "source": "deploy.md" },
+        { "text": "CI/CD pipeline setup guide", "score": 0.82, "source": "ci.md" }
+      ],
+      "resultCount": 4
+    }
+  },
+  "expect": {
+    "steps": {
+      "search_1": { "status": "completed" },
+      "search_2": { "status": "completed" },
+      "merge": { "status": "completed" },
+      "rerank_all": { "status": "completed" }
+    },
+    "output": {
+      "results": { "type": "array", "minLength": 1 },
+      "resultCount": { "type": "number" }
+    },
+    "noErrors": true
+  }
+}

package/src/workflows/tests/multi-collection-search.one-empty.test.json

@@ -0,0 +1,28 @@
+{
+  "name": "multi-collection-search: one collection empty",
+  "inputs": {
+    "query": "obscure feature",
+    "collection1": "api_docs",
+    "collection2": "tutorials",
+    "limit": 5
+  },
+  "mocks": {
+    "query": {
+      "results": [],
+      "resultCount": 0
+    },
+    "rerank": {
+      "results": [],
+      "resultCount": 0
+    }
+  },
+  "expect": {
+    "steps": {
+      "search_1": { "status": "completed" },
+      "search_2": { "status": "completed" },
+      "merge": { "status": "completed" },
+      "rerank_all": { "status": "completed" }
+    },
+    "noErrors": true
+  }
+}

package/src/workflows/tests/rag-chat.happy-path.test.json

@@ -0,0 +1,26 @@
+{
+  "name": "rag-chat: happy path with relevant results",
+  "inputs": {
+    "question": "How do I create a vector search index?",
+    "collection": "docs",
+    "collection2": "",
+    "limit": 5,
+    "min_score": 0.3,
+    "system_prompt": "You are a helpful assistant. Cite sources.",
+    "chat_history": ""
+  },
+  "mocks": {
+    "query": { "results": [{ "text": "To create a vector search index, use the Atlas UI or CLI...", "source": "atlas-docs.md", "score": 0.92 }, { "text": "Vector indexes support cosine, euclidean, and dotProduct similarity.", "source": "vector-guide.md", "score": 0.85 }], "resultCount": 2 },
+    "rerank": { "results": [{ "text": "To create a vector search index, use the Atlas UI or CLI...", "source": "atlas-docs.md", "score": 0.95 }, { "text": "Vector indexes support cosine, euclidean, and dotProduct similarity.", "source": "vector-guide.md", "score": 0.78 }], "resultCount": 2 },
+    "generate": { "text": "To create a vector search index, you can use the Atlas UI or the CLI. According to the documentation (atlas-docs.md), the process involves...", "model": "claude-sonnet", "provider": "anthropic" }
+  },
+  "expect": {
+    "steps": {
+      "search_primary": { "status": "completed" },
+      "rerank": { "status": "completed" },
+      "filter_relevant": { "status": "completed" },
+      "generate_answer": { "status": "completed" }
+    },
+    "noErrors": true
+  }
+}

package/src/workflows/tests/rag-chat.no-relevant-results.test.json

@@ -0,0 +1,25 @@
+{
+  "name": "rag-chat: no results above relevance threshold",
+  "inputs": {
+    "question": "What is the meaning of life?",
+    "collection": "technical-docs",
+    "collection2": "",
+    "limit": 5,
+    "min_score": 0.5,
+    "system_prompt": "You are a helpful assistant.",
+    "chat_history": ""
+  },
+  "mocks": {
+    "query": { "results": [{ "text": "MongoDB Atlas provides cloud database services.", "source": "overview.md", "score": 0.15 }], "resultCount": 1 },
+    "rerank": { "results": [{ "text": "MongoDB Atlas provides cloud database services.", "source": "overview.md", "score": 0.12 }], "resultCount": 1 }
+  },
+  "expect": {
+    "steps": {
+      "search_primary": { "status": "completed" },
+      "rerank": { "status": "completed" },
+      "filter_relevant": { "status": "completed" },
+      "no_context_response": { "status": "completed" }
+    },
+    "noErrors": true
+  }
+}

package/src/workflows/tests/research-and-summarize.happy-path.test.json

@@ -0,0 +1,33 @@
+{
+  "name": "research-and-summarize: happy path",
+  "inputs": {
+    "question": "How does vector search work?",
+    "limit": 5
+  },
+  "mocks": {
+    "query": {
+      "results": [
+        { "text": "Vector search uses embeddings to find similar documents", "score": 0.95 },
+        { "text": "Approximate nearest neighbor algorithms power vector search", "score": 0.88 }
+      ],
+      "resultCount": 2
+    },
+    "generate": {
+      "text": "Vector search works by converting documents and queries into numerical embeddings and finding the closest matches using ANN algorithms.",
+      "model": "claude-3-haiku",
+      "tokensUsed": 120
+    }
+  },
+  "expect": {
+    "steps": {
+      "research": { "status": "completed" },
+      "summarize": { "status": "completed" }
+    },
+    "output": {
+      "summary": { "type": "string", "minLength": 1 },
+      "sources": { "type": "array", "minLength": 1 },
+      "sourceCount": { "type": "number" }
+    },
+    "noErrors": true
+  }
+}

package/src/workflows/tests/research-and-summarize.no-results.test.json

@@ -0,0 +1,29 @@
+{
+  "name": "research-and-summarize: no search results",
+  "inputs": {
+    "question": "What is the meaning of life?",
+    "limit": 5
+  },
+  "mocks": {
+    "query": {
+      "results": [],
+      "resultCount": 0
+    },
+    "generate": {
+      "text": "No relevant documents were found in the knowledge base to answer this question.",
+      "model": "claude-3-haiku",
+      "tokensUsed": 45
+    }
+  },
+  "expect": {
+    "steps": {
+      "research": { "status": "completed" },
+      "summarize": { "status": "completed" }
+    },
+    "output": {
+      "summary": { "type": "string", "minLength": 1 },
+      "sourceCount": { "type": "number" }
+    },
+    "noErrors": true
+  }
+}

package/src/workflows/tests/search-with-fallback.empty-both.test.json

@@ -0,0 +1,24 @@
+{
+  "name": "search-with-fallback: both collections empty",
+  "inputs": {
+    "query": "nonexistent topic xyz",
+    "primary_collection": "api_docs",
+    "fallback_collection": "knowledge"
+  },
+  "mocks": {
+    "query": {
+      "results": [],
+      "resultCount": 0
+    }
+  },
+  "expect": {
+    "steps": {
+      "primary_search": { "status": "completed" },
+      "check_results": { "status": "completed" },
+      "format_primary": { "status": "skipped" },
+      "fallback_search": { "status": "completed" },
+      "format_fallback": { "status": "completed" }
+    },
+    "noErrors": true
+  }
+}

package/src/workflows/tests/search-with-fallback.fallback-branch.test.json

@@ -0,0 +1,24 @@
+{
+  "name": "search-with-fallback: fallback branch (primary empty)",
+  "inputs": {
+    "query": "obscure topic",
+    "primary_collection": "api_docs",
+    "fallback_collection": "knowledge"
+  },
+  "mocks": {
+    "query": {
+      "results": [],
+      "resultCount": 0
+    }
+  },
+  "expect": {
+    "steps": {
+      "primary_search": { "status": "completed" },
+      "check_results": { "status": "completed" },
+      "format_primary": { "status": "skipped" },
+      "fallback_search": { "status": "completed" },
+      "format_fallback": { "status": "completed" }
+    },
+    "noErrors": true
+  }
+}

package/src/workflows/tests/search-with-fallback.happy-path.test.json

@@ -0,0 +1,27 @@
+{
+  "name": "search-with-fallback: happy path (primary has results)",
+  "inputs": {
+    "query": "how to authenticate",
+    "primary_collection": "api_docs",
+    "fallback_collection": "knowledge"
+  },
+  "mocks": {
+    "query": {
+      "results": [
+        { "text": "Use Bearer tokens for API authentication", "score": 0.95, "source": "auth.md" },
+        { "text": "OAuth2 is supported for third-party apps", "score": 0.88, "source": "oauth.md" }
+      ],
+      "resultCount": 2
+    }
+  },
+  "expect": {
+    "steps": {
+      "primary_search": { "status": "completed" },
+      "check_results": { "status": "completed" },
+      "format_primary": { "status": "completed" },
+      "fallback_search": { "status": "skipped" },
+      "format_fallback": { "status": "skipped" }
+    },
+    "noErrors": true
+  }
+}

package/src/workflows/tests/smart-ingest.duplicate-detected.test.json

@@ -0,0 +1,34 @@
+{
+  "name": "smart-ingest: duplicate detected (high similarity)",
+  "description": "When a near-duplicate is found (similarity above 0.85), the ingest_doc step is correctly skipped to avoid ingesting duplicate content.",
+  "inputs": {
+    "text": "A document that already exists in the knowledge base",
+    "source": "existing-doc.md",
+    "threshold": 0.85
+  },
+  "mocks": {
+    "search": {
+      "results": [
+        { "text": "A document that already exists in the knowledge base", "score": 0.98 }
+      ],
+      "resultCount": 1
+    },
+    "similarity": {
+      "similarity": 0.97,
+      "model": "voyage-4-large"
+    },
+    "ingest": {
+      "chunks": 1,
+      "source": "existing-doc.md",
+      "collection": "default"
+    }
+  },
+  "expect": {
+    "steps": {
+      "check_existing": { "status": "completed" },
+      "similarity_check": { "status": "completed" },
+      "ingest_doc": { "status": "skipped" }
+    },
+    "noErrors": true
+  }
+}

package/src/workflows/tests/smart-ingest.happy-path.test.json

@@ -0,0 +1,31 @@
+{
+  "name": "smart-ingest: novel document (no similar docs found)",
+  "inputs": {
+    "text": "A completely new document about quantum computing fundamentals",
+    "source": "quantum-101.md",
+    "threshold": 0.85
+  },
+  "mocks": {
+    "search": {
+      "results": [],
+      "resultCount": 0
+    },
+    "similarity": {
+      "similarity": 0.12,
+      "model": "voyage-4-large"
+    },
+    "ingest": {
+      "chunks": 3,
+      "source": "quantum-101.md",
+      "collection": "default"
+    }
+  },
+  "expect": {
+    "steps": {
+      "check_existing": { "status": "completed" },
+      "similarity_check": { "status": "skipped" },
+      "ingest_doc": { "status": "completed" }
+    },
+    "noErrors": true
+  }
+}