conversation-search

Search indexed conversation history with SQLite full-text search.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/dhughes/claude-marketplace --skill conversation-search-dhughes
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: conversation-search
Source: https://github.com/dhughes/claude-marketplace/tree/main/plugins/conversation-index/skills/conversation-search
Command: npx skills add https://github.com/dhughes/claude-marketplace --skill conversation-search-dhughes

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires better-sqlite3, and includes scripts (resource) components.

What problem does it solve?

This Skill lets you search indexed Claude Code conversations to find past discussions, identify topics, and determine which projects conversations occurred in.

Core Features & Use Cases

  • Full-text search across conversations using SQLite FTS5 for fast results.
  • Supports scope: current project or all projects.
  • Returns: total matches, per-conversation summaries, and relevance scores.
  • Use Case: Quickly answer "when did we discuss X?" or "which projects mention Y?" across all projects.

Quick Start

To search conversations, run the script with a query, e.g.: cd <skill-base-directory>/scripts && ./search.sh --json --scope current_project "authentication"

Frequently Asked Questions about conversation-search

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I search across Claude Code conversations to find past discussions?

Search indexed conversation history using SQLite full-text search with the search script. Run `./search.sh --json --scope current_project "query"` to find conversations by topic, return match counts, per-conversation summaries, and relevance scores across projects or within your current project.

Can I search conversations across multiple projects at once?

Yes. The search scope option supports both current_project and all_projects modes, letting you query indexed conversations across all projects simultaneously to identify which projects discussed a specific topic.

What does SQLite full-text search do for conversation indexing?

SQLite FTS5 indexes conversation content for fast full-text queries, enabling rapid pattern matching across large conversation histories without scanning raw files, returning ranked results with relevance scores.

Do I need to index conversations before searching?

Conversations must be indexed first using the Skill's indexing process before search queries return results. The indexed data is stored in SQLite, enabling efficient FTS5 queries across projects.

What output formats does conversation search support?

Search returns JSON or human-readable formats via the --json flag, providing total match counts, per-conversation summaries with context, and relevance scores for each result.

When should I use conversation search instead of manual review?

Use search when you need to answer "when did we discuss X?" or "which projects mention Y?" across many conversations—it reduces manual scanning and identifies relevant discussions with relevance ranking.