Import

HAEVN can import conversations from external sources — useful for seeding your archive with existing exports before you start using the sync extension, or for restoring from a backup.

Supported import formats

Source	Format	How to get it
ChatGPT	ZIP file	Settings → Data controls → Export data
Claude Code	JSONL file	Local session files in your `.claude/` directory
Codex	JSONL file	Local session files in your `.codex/sessions/` directory
PI	JSONL file	Local session files in your `.pi/sessions/` directory
HAEVN backup	ZIP	From a previous HAEVN export
HAEVN Markdown	`.md`	YAML-frontmatter markdown transcript

Open the archive (click Archive in the HAEVN popup, or right-click the extension icon → Options).
Click the IMPORT button in the archive toolbar.
In the import dialog, click Choose file and select your ZIP or JSON export file.
Click Import. HAEVN processes the file and adds the conversations to your archive.
The search index rebuilds automatically once the import completes.

# Claude Code
haevn import --format claude_code ~/.claude/projects/my-proj/session.jsonl

# Codex
haevn import --format codex ~/.codex/sessions/2026/03/23/rollout-...jsonl

# PI
haevn import --format pi ~/.pi/sessions/2026-03-23T14-48-43-844Z_...jsonl

# HAEVN Markdown
haevn import --format markdown ./conversation.md

Useful flags: --no-overwrite to skip existing chat IDs, and --skip-index to skip index rebuild after import.

Importing from ChatGPT

ChatGPT’s native export format is a ZIP file containing your full conversation history. To export from ChatGPT:

Go to chatgpt.com and open Settings.
Navigate to Data controls → Export data.
Click Export and wait for the email with your download link (this can take several minutes to hours depending on your history size).
Download the ZIP file from the link in the email.
Import the ZIP file into HAEVN using the steps above.

For the exact HAEVN Markdown schema (YAML frontmatter + role delimiters), see Reference → Data Formats.

What happens during import

Each conversation is added to your archive with its original timestamps and metadata
Attachments and images are extracted and stored in OPFS
Existing chats are overwritten by default (use --no-overwrite in CLI to skip existing IDs)
The search index is rebuilt after all conversations are imported (unless --skip-index in CLI)