Save a new conversation context for cross-agent sharing. Contexts are stored and indexed in a vector store for semantic search. Supports organization-level contexts for team collaboration.
API key must be provided in the Authorization header
Descriptive title for the context
1 - 200Brief summary of the conversation
10 - 1000Full conversation context
50Which agent is creating this context (e.g., "cursor", "claude-code")
Searchable tags for categorization
Optional metadata like file paths, repositories discussed, etc.
References to NIA resources used during the conversation
List of files that were modified during conversation
Optional custom workspace name (overrides auto-detection)
Memory taxonomy type:
scratchpad, episodic, fact, procedural Custom TTL in seconds. If not set, uses memory_type default (scratchpad=3600, episodic=604800, fact/procedural=permanent)
x >= 0Provenance tracking input for content artifacts
Context saved successfully
Unique identifier for the context
User who created the context
Organization ID if context belongs to an organization
Context title
Context summary
Full context content
Source agent (e.g., "cursor", "claude-code")
When the context was created
When the context was last updated
References to NIA resources used during the conversation
Memory taxonomy type:
scratchpad, episodic, fact, procedural When this context will expire (null for permanent memory types)
Provenance tracking metadata for content artifacts