search

name	search
description	Semantic search across all past conversations, tool results, AI outputs, and sub-agent activity stored in the RAG database
disable-model-invocation	true
allowed-tools	Bash(curl *)
argument-hint	[query]

RAG Search

Search the RAG database for content matching the user's query: "$ARGUMENTS"

Instructions

1. Read the plugin config to get the backend endpoint:

   cat ~/.claude/plugins/claude-rag/config.json 2>/dev/null || echo '{"connection":{"endpoint":"https://api.clauderag.io"}}'
   

2. Execute the search query against the backend:

   curl -s -X POST <endpoint>/api/v1/search \
     -H "Content-Type: application/json" \
     -d '{"query": "<user_query>", "limit": 10, "threshold": 0.5}'
   

3. Display results in a clear, organized format:

   • Group by relevance score (high to low)
   • For each result show:
     • Content type icon: Prompt=💬, AI Response=🤖, Tool Result=🔧, Error=❌
     • Score (as percentage)
     • Tool name (if applicable)
     • Agent type + sub-agent indicator
     • Project name
     • Date
     • Content preview (first 200 chars)
   • If no results: suggest broadening the query or checking filters

4. If the user provides filters in their query, parse them:

   • "in project X" → {"filters": {"project_id": "X"}}
   • "from sub-agents" → {"filters": {"is_sub_agent": true}}
   • "only tools" → {"filters": {"content_types": ["tool_call", "tool_result"]}}
   • "bash results" → {"filters": {"tool_names": ["Bash"]}}
   • "code files" → {"filters": {"tool_categories": ["file"]}}

5. Show total results count and search latency at the bottom.