[FEATURE] Implement Pinecone Vector Store

[Oxazolidinone]

Add Pinecone as Vector Store Backend

Problem

Cipher's current architecture for vector storage is limited to Qdrant and Milvus. This constraint prevents users who rely on Pinecone for their cloud-based vector search and management needs from integrating their existing infrastructure with Cipher.

Solution

Add Pinecone as a supported vector store backend following the existing architecture patterns.

Implement

User Experience

Users will be able to configure Cipher to connect to Pinecone using the following environment variables:

• VECTOR_STORE_TYPE=pinecone
• PINECONE_API_KEY=<your-pinecone-api-key>
• PINECONE_ENVIRONMENT=<your-pinecone-environment>
• PINECONE_INDEX=knowledge-memory
• PINECONE_NAMESPACE=default (optional)
• VECTOR_STORE_DIMENSION=1536

Technical Changes

• Add dependency: Include @pinecone-database/pinecone in package.json.
• Create backend: Develop src/core/vector_storage/backend/pinecone.ts.
• Update factory: Extend the getVectorStoreConfigFromEnv() function in factory.ts to support Pinecone's configuration.
• Update types: Integrate Pinecone into the existing vector store type system.
• Update docs: Add setup instructions to /docs/vector-stores.md.

Requirements

• Implement the VectorStore interface from backend/types.ts.
• Fully support Pinecone features such as namespaces and metadata filtering.
• Handle authentication via the API key (PINECONE_API_KEY).
• Include proper error handling, connection management, and rate limiting.
• Add comprehensive test coverage (unit + integration).
• Support workspace memory and multi-collection features by leveraging Pinecone's namespaces.

Acceptance Criteria

• Pinecone client dependency added.
• Pinecone backend fully implemented.
• Environment variable configuration supported.
• Factory method updated to recognize Pinecone.
• Test coverage (unit and integration) achieved.
• Documentation updates
• Pinecone namespaces are supported.
• Workspace memory collection supported.

Priority

• Priority: Medium
• Impact: Medium
• Effort: Medium