Prompt Library

Store this content as a memory in Trix:
1. Connect to the Trix API using my API key
2. Select the target space
3. Store the memory with appropriate metadata
4. Verify the memory was stored successfully
5. Enrich the memory to extract entities and topics

Project context:
- SDK: [Python/TypeScript/C#/REST API]
- Space: [your space name or ID]
- Content type: [conversation/document/note/meeting notes]

Batch store multiple memories into Trix:
- Up to 100 memories per batch (API limit)
- Each memory has: content, optional metadata, optional tags
- Handle partial failures gracefully
- Log which memories succeeded and which failed
- Use the batch_store_memories endpoint

Source data:
- Format: [JSON array/CSV/text files]
- Count: [number of memories]
- Space: [target space]

Help me manage memories in my Trix space:
1. Search for memories matching [criteria]
2. Update memory content or metadata
3. Pin important memories so they persist
4. Soft-delete outdated memories
5. Restore accidentally deleted memories

Space: [space name or ID]
Operation: [update/pin/delete/restore]
Target: [memory ID or search criteria]

Check the quality of memories in my Trix space:
1. Get memory health metrics for the space
2. Identify low-quality or duplicate memories
3. Find memories missing enrichment data
4. Get topic distribution across memories
5. Suggest improvements for memory organization

Space: [space name or ID]
Focus: [quality/duplicates/enrichment/topics]

Search my Trix memories using semantic search:
1. Use the semantic_search endpoint for meaning-based matching
2. Return the top [N] most relevant results
3. Include relevance scores in the output
4. Filter by date range: [start] to [end]
5. Filter by tags: [tag list]

Query: [your natural language search query]
Space: [space name or ID]

Perform a hybrid search combining semantic and keyword matching:
1. Use fulltext_search for exact keyword matches
2. Use semantic_search for meaning-based results
3. Combine and deduplicate results
4. Rank by combined relevance
5. Return structured results with highlights

Query: [search query]
Keywords: [specific terms to match]
Space: [space name or ID]

Run a deep search across my Trix memories:
1. Use deep_search for comprehensive multi-pass retrieval
2. Include search explanation showing why results matched
3. Generate related queries for follow-up exploration
4. Aggregate results across multiple spaces if needed
5. Summarize the key findings from matched memories

Query: [complex search query]
Spaces: [space IDs, or "all"]
Explain: true

Find memories related to a specific memory:
1. Start with memory ID: [memory ID]
2. Use find_related_memories to get connected content
3. Also check the knowledge graph for entity relationships
4. Show how each related memory connects
5. Suggest clusters or topics that link them

Space: [space name or ID]
Depth: [1-3 hops]

Explore the knowledge graph in my Trix space:
1. List all entities extracted from my memories
2. Get the subgraph for entity: [entity name]
3. Show relationships between entities
4. Find the most connected entities (hubs)
5. Identify isolated entities with few connections

Space: [space name or ID]
Focus entity: [name or "overview"]

Work with relationships in the Trix knowledge graph:
1. Create a relationship between [entity A] and [entity B]
2. Set relationship type: [type]
3. List existing relationships for [entity]
4. Find the shortest path between two entities
5. Suggest new relationships based on memory content

Space: [space name or ID]
Relationship type: [related_to/part_of/caused_by/depends_on/etc.]

Find contradictions in my Trix memories:
1. Scan for conflicts across the space
2. List detected contradictions with evidence
3. Show the conflicting memories side by side
4. Suggest which version is more likely correct
5. Resolve or dismiss conflicts as appropriate

Space: [space name or ID]
Scope: [all/recent/specific topic]

Work with facts extracted from my memories:
1. List facts for a given memory or entity
2. Create new facts from structured data
3. Find facts that support or contradict a claim
4. Get fact-based answers to questions
5. Verify fact accuracy against source memories

Space: [space name or ID]
Claim: [statement to verify]

Set up the Trix Python SDK in my project:
1. Install trixdb via pip
2. Configure the async client with API key and base URL
3. Initialize with proper error handling
4. Store a test memory and verify it works
5. Implement retry logic for transient failures
6. Add type hints throughout

Requirements:
- Python version: [3.10+]
- Async: [yes/no]
- Framework: [FastAPI/Django/Flask/standalone]

Set up the Trix TypeScript SDK in my project:
1. Install @trixdb/client via npm
2. Configure the client with API key and base URL
3. Add proper TypeScript types for responses
4. Store and search memories with full type safety
5. Handle errors with typed error responses
6. Configure retry and timeout options

Requirements:
- Runtime: [Node.js/Bun/Deno]
- Framework: [Express/Next.js/standalone]
- TypeScript version: [5.x]

Set up the Trix C# SDK (Trix NuGet package):
1. Install Trix via NuGet
2. Configure TrixClient with API key
3. Implement dependency injection registration
4. Store and retrieve memories
5. Handle exceptions and rate limiting
6. Add logging for API calls

Requirements:
- .NET version: [8.0/9.0]
- Project type: [ASP.NET/Console/Worker Service]

Set up a webhook listener for Trix events:
1. Create an HTTP endpoint to receive Trix webhooks
2. Validate webhook signatures for security
3. Handle events: memory.created, memory.updated, memory.deleted
4. Process enrichment completion events
5. Implement idempotency to handle duplicate deliveries
6. Add error handling and retry logic

Framework: [Express/FastAPI/ASP.NET/etc.]
Events: [list which events to handle]

Help me set up Trix MCP server with Claude Desktop:
1. Configure the MCP server in Claude Desktop settings
2. Set the API key and base URL environment variables
3. Verify the connection works with a test query
4. List available tools to confirm 140+ tools are loaded
5. Store a test memory using the MCP tools

Environment:
- OS: [macOS/Windows/Linux]
- Claude Desktop version: [version]
- Trix API key: [configured in env]

Using Trix MCP tools, store these meeting notes as memories:
1. Use store_memory to save the meeting content
2. Add metadata: date, participants, topic
3. Tag with relevant categories
4. Enrich the memory to extract entities and topics
5. Link to related previous meeting memories
6. Create tasks extracted from action items

Meeting content:
[paste your meeting notes]
Space: [space name or ID]

Using Trix MCP tools, research everything I know about [topic]:
1. Search memories about [topic] using semantic search
2. Find related entities in the knowledge graph
3. Get topic-specific memories using search_by_topic
4. Summarize findings across matched memories
5. Identify knowledge gaps on this topic
6. Generate follow-up questions for deeper exploration

Topic: [your topic]
Space: [space name or ID, or "all"]

Set up a daily knowledge management workflow with Trix MCP:
1. Create a daily note with today's date
2. Review due habits and check in on completed ones
3. Check goal progress and update key results
4. Search for memories related to today's tasks
5. Store new learnings from the day as memories
6. Review and resolve any detected conflicts

Persona: [if using personas]
Space: [space name or ID]