Memory Status Command

Check the indexing status and health of Infinite Memory for the current project.

Your Task

1. Get the current working directory:

   pwd
   

2. Check indexing status by querying the database:

   • Use query_memory MCP tool with current directory
   • Try a simple test query (e.g., "database" or "API")
   • If results found → Project is indexed
   • If no results → Project not indexed or needs indexing

3. Present status in clear format:

   If Indexed (query returned results):

   📊 Memory Status: /path/to/project
   
   ✅ Indexed - {N} results found for test query
   
   Your code is searchable! Use:
   - /infinite-memory:search <query> - Semantic search
   - /infinite-memory:index-project - Update index
   - /infinite-memory:re-index - Force full re-index
   

   If Not Indexed (query returned 0 results):

   ❌ Not Indexed
   
   This project hasn't been indexed yet.
   
   To enable semantic search, run: /infinite-memory:index-project
   
   This will:
   - Index all supported files (.py, .js, .ts, .md, etc.)
   - Generate embeddings for semantic search
   - Enable natural language code queries
   

Example Outputs

Project Indexed

📊 Memory Status: /Users/dev/projects/api-service

✅ Indexed - 5 results found for "API"

Your code is searchable! Use:
- /infinite-memory:search <query> - Semantic search
- /infinite-memory:index-project - Update index
- /infinite-memory:re-index - Force full re-index

Project Not Indexed

❌ Not Indexed

This project hasn't been indexed yet.

To enable semantic search, run: /infinite-memory:index-project

This will:
- Index all supported files (.py, .js, .ts, .md, etc.)
- Generate embeddings for semantic search
- Enable natural language code queries

Troubleshooting Suggestions

If status shows issues:

No files indexed:

• Check permissions on project directory
• Ensure project contains supported file types
• Try: /index-project again

Many files failed:

• Check error logs
• Verify file permissions
• Ensure not indexing system directories

Slow searches:

• Check if database is on SSD (HDD will be slower)
• Ensure at least 4GB RAM available
• Consider excluding large directories

Implementation Notes

• Use only 2 operations: pwd + query_memory (with max_results: 1)
• Don't run bash commands like shasum, ls, du, stat
• Don't calculate project hashes manually - the MCP tool handles this
• Keep it fast (<2 seconds) by avoiding unnecessary filesystem operations
• The query result count tells us if the project is indexed