A utility to migrate ChatGPT conversation history into OpenClaw's memory search by converting exports into searchable SQLite vector databases.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install chatgpt-import
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install chatgpt-import using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is ChatGPT History Import?
The chatgpt-import skill allows users to bridge the gap between their legacy ChatGPT interactions and their new local AI workflows. By importing historical data, it ensures that your AI agent has access to past research, code snippets, and creative sessions. This process is a vital component for power users of Openclaw Skills who want to maintain a persistent and searchable knowledge base across different platforms.
This skill automates the transformation of ChatGPT's proprietary JSON export into structured Markdown and subsequently into an embedded SQLite database. This enables the OpenClaw engine to retrieve context from years of previous conversations, making your local agent significantly more informed and personalized.
ChatGPT History Import Use Cases
- Migrating historical research and data from ChatGPT to a local-first OpenClaw environment.
- Building a searchable archive of past AI interactions for personal knowledge management.
- Enhancing OpenClaw context by providing it with years of specific user-AI interaction history.
- Centralizing disparate AI conversation logs into the unified Openclaw Skills memory architecture.
How ChatGPT History Import Works
- The user requests a data export from ChatGPT to receive a comprehensive conversations.json file.
- A specialized Python script parses the JSON export and converts each conversation into an individual Markdown file.
- A bulk embedding script processes these files, sending text chunks to an embedding API to generate vector representations.
- The resulting vectors and text content are stored in a local SQLite database designed for high-speed similarity search.
- The OpenClaw configuration is updated to include the new SQLite database as an additional memory search path.
- Upon restarting the gateway, the historical data becomes accessible to Openclaw Skills queries.
ChatGPT History Import Setup
First, follow the standard ChatGPT export procedure to obtain your data. Once you have the data, convert it to Markdown:
python3 scripts/convert_chatgpt.py \
--input /path/to/conversations.json \
--output /path/to/chatgpt-history
Next, generate the embeddings and create the SQLite database:
export OPENAI_API_KEY=sk-...
python3 scripts/bulk_embed.py \
--history-dir /path/to/chatgpt-history \
--db /path/to/chatgpt-memory.sqlite
Finally, add the database path to your config.yaml and restart the gateway:
openclaw gateway restart
ChatGPT History Import Data Schema & Taxonomy
The skill manages data across several stages to ensure compatibility with Openclaw Skills memory search:
| Component | Format | Description |
|---|---|---|
| Raw Export | JSON | The original data packet from ChatGPT containing all metadata. |
| Archive Store | Markdown | Individual files converted from JSON, filtered by message count. |
| Vector Memory | SQLite | The final searchable database containing text chunks and their corresponding embeddings. |
| Configuration | YAML | The OpenClaw settings file that points to the local SQLite storage. |
ChatGPT History Import Advanced Features
- Message filtering via
--min-messagesto exclude short or low-value interactions from the index. - High-concurrency embedding using
--max-workersto speed up the processing of large conversation histories. - Flexible embedding model support, allowing users to balance cost and search accuracy (e.g., text-embedding-3-small).
- Support for multiple memory paths, allowing ChatGPT history to coexist with other Openclaw Skills data sources.
SKILL.md
Loading
Related Openclaw Skills
MemePickup Wingman
An AI-powered dating assistant that generates personalized openers, conversation replies, and profile analysis using the MemePickup API.
n8nops
An expert-level automation agent for building, debugging, and monitoring n8n workflows through natural language and the n8n REST API.
Intros
A specialized social network integrated into your AI agent that facilitates professional discovery, connection management, and secure messaging.
OEJTS Personality Tuner
A sophisticated tool for administering OEJTS assessments and tuning AI agent behavior based on validated personality results.
Oura Ring Integration
A Python-powered skill for fetching Oura Ring sleep, readiness, and resilience data through the Oura V2 API.
Reader Deep Dive
A sophisticated intelligence briefing skill that bridges the gap between your current reading list and your historical knowledge archive.
Featured*
