From episodic-memory
Searches conversation history for approaches after code exploration, solutions when stuck, unfamiliar workflows, or past references. Dispatches search-conversations agent.
npx claudepluginhub obra/episodic-memory --plugin episodic-memoryThis skill uses the workspace's default tool permissions.
**Core principle:** Search before reinventing. Searching costs nothing; reinventing or repeating mistakes costs everything.
Recalls, searches, and analyzes past conversations using recent_chats.py, search_conversations.py, and lenses like retro, find-gaps, extract-decisions for context restoration and retrospectives.
Semantically searches past Claude Code sessions to recall commands, solutions, and context from prior conversations. Triggers on 'remember when', 'last time', or /woz-recall.
Searches Claude Code conversation history in Scribe DB (SQLite FTS5) or JSONL files to recall past discussions, decisions, code snippets, and context from prior sessions.
Share bugs, ideas, or general feedback.
Core principle: Search before reinventing. Searching costs nothing; reinventing or repeating mistakes costs everything.
YOU MUST dispatch the search-conversations agent for any historical search.
Announce: "Dispatching search agent to find [topic]."
Then use the Task tool with subagent_type: "search-conversations":
Task tool:
description: "Search past conversations for [topic]"
prompt: "Search for [specific query or topic]. Focus on [what you're looking for - e.g., decisions, patterns, gotchas, code examples]."
subagent_type: "search-conversations"
The agent will:
search toolshow toolSaves 50-100x context vs. loading raw conversations.
You often get value out of consulting your episodic memory once you understand what you're being asked. Search memory in these situations:
After understanding the task:
When you're stuck:
When historical signals are present:
Don't search first:
You CAN use MCP tools directly, but DON'T:
mcp__plugin_episodic-memory_episodic-memory__searchmcp__plugin_episodic-memory_episodic-memory__showUsing these directly wastes your context window. Always dispatch the agent instead.
See MCP-TOOLS.md for complete API reference if needed for advanced usage.