aii-chat-history-search

Indexes AI chat exports into a local SQLite FTS5 database for full-text search and thread retrieval.

Updated Aug 21, 2026
One-click install
npx skills add https://github.com/maestroventures/aiintegrator --skill aii-chat-history-search-maestroventures
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: aii-chat-history-search
Source: https://github.com/maestroventures/aiintegrator/tree/main/plugins/aii-blueprint/skills/aii-chat-history-search
Command: npx skills add https://github.com/maestroventures/aiintegrator --skill aii-chat-history-search-maestroventures

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? AI chat exports from Claude or ChatGPT accumulate years of decisions, wording, and solved problems in a JSON file that cannot be searched. This Skill converts that export into a local, full-text searchable knowledge base and distills a reusable thinking profile from the user's strongest threads, with all data staying on the user's own machine. ## Core Features & Use Cases - Local Full-Text Search: Builds a SQLite + FTS5 index from Claude, ChatGPT, or generic JSONL exports and supports ranked snippet search with phrases, AND/OR/NOT, and prefix queries. - Thread Dumping and Listing: Retrieves any conversation verbatim by ID, lists the largest threads, and reports database statistics for shape checks. - Thinking Profile Distillation: Extracts verbatim reasoning from the strongest threads to produce an auditable thinking-profile.md capturing how the user decides and writes. - Use Case: A consultant asks "where did I decide on blended pricing?" — the Skill runs an FTS query against their local index, returns ranked snippets, and dumps the full thread so they can lift the exact wording that worked. ## Quick Start Ask the assistant to build a searchable local index from your Claude or ChatGPT conversations.json export and then search it for a past decision or topic.

Frequently Asked Questions about aii-chat-history-search

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

FAQPage Schema
How do I search my Claude or ChatGPT chat history?

Export your conversations.json from Claude (Settings → Account → Export data) or ChatGPT (Settings → Data controls → Export), then run build_index.py to create a local SQLite FTS5 database. Use query.py with FTS5 syntax for ranked snippet search or thread dumps.

What export formats does the chat history indexer support?

The indexer auto-detects Claude exports (conversations.json with chat_messages) and ChatGPT exports (conversations.json with mapping nodes). It also accepts a generic JSONL format with one object per line containing conversation_id, title, role, text, and created_at fields.

Does my chat export get uploaded anywhere when building the search index?

No, the export and database never leave your machine. The build and query scripts run locally via Python, and the AI session only writes the commands and reads back results you choose to share.

What query syntax does the FTS5 chat search support?

The search uses SQLite FTS5 syntax: bare words, double-quoted exact phrases, AND/OR/NOT boolean operators, and prefix matching with an asterisk. Results return ranked by BM25 with highlighted snippets, conversation title, role, and ID.

Why does the index build fail with a zero messages error?

The builder fails loudly when the export parses but contains no message text, rather than reporting success on an empty index. Check that the export file is a valid Claude or ChatGPT conversations.json, or pass --format jsonl for generic line-delimited input.

Can I keep chat indexes for multiple clients separate?

Yes, name each database with the client's registered key, such as chat_history-<client-key>.db. Rebuilding with the same client's export is idempotent, but building a different client's export against the same filename silently replaces the previous index.