remember

/remember Command

Save the current session to context memory with an optional annotation.

Usage

/remember [note]

Arguments:

• note (optional): A personal annotation or tag to help find this session later

Examples

/remember
/remember "Fixed the auth bug with refresh tokens"
/remember "Important: OAuth2 implementation details"

Workflow

When the user runs /remember:

1. Generate Session Summary

   Analyze the current conversation and create:

   • brief: A single sentence summarizing what was accomplished
   • detailed: 2-3 paragraphs with full context of what happened
   • key_decisions: List of important decisions made during the session
   • problems_solved: List of problems that were resolved
   • technologies: List of technologies, frameworks, or tools discussed
   • outcome: One of: success, partial, abandoned

2. Extract Topics

   Identify 3-8 relevant topics from the conversation. Use lowercase, common terms like:

   • Technology names: react, python, sqlite
   • Concepts: authentication, debugging, refactoring
   • Domains: api, frontend, database

3. Identify Key Code

   If significant code was written or discussed, extract important snippets with:

   • The code itself
   • The programming language
   • A brief description of what it does
   • The file path if applicable

4. Extract Key Messages

   Select 5-15 important messages from the conversation that capture:

   • The initial request/problem statement
   • Key decisions and their reasoning
   • Solution descriptions
   • Important caveats or warnings

5. Pipe JSON via Stdin and Save to Database

   Pipe JSON directly via --json - (stdin):

   python "~/.claude/skills/context-memory/scripts/db_save.py" --json - << 'ENDJSON'
   {
     "session_id": "<UNIQUE_ID>",
     "project_path": "<PROJECT_PATH>",
     "messages": [
       {"role": "user", "content": "The initial question or request"},
       {"role": "assistant", "content": "The response or solution"}
     ],
     "summary": {
       "brief": "One-line summary of what was accomplished",
       "detailed": "2-3 paragraphs with full context...",
       "key_decisions": ["Decision 1", "Decision 2"],
       "problems_solved": ["Problem 1", "Problem 2"],
       "technologies": ["python", "sqlite", "fts5"],
       "outcome": "success"
     },
     "topics": ["topic1", "topic2", "topic3"],
     "code_snippets": [
       {
         "code": "def example(): pass",
         "language": "python",
         "description": "What this code does",
         "file_path": "src/example.py"
       }
     ],
     "user_note": "User's note if provided, or null"
   }
   ENDJSON