Deepchat Data Import

When you're building a tool that needs to read DeepChat's local databases, this skill walks you through the entire import path: finding agent.db on disk, handling SQLCipher encryption and Electron safeStorage wrappers, navigating the SQLite schema for providers, sessions, messages, and legacy chat.db migrations. The safety rules are the real value here. They're explicit about consent, warn you not to write to a live profile, and remind you to copy WAL files or wait for the app to quit. References cover data locations across macOS, Windows, and Linux, plus schema inspection strategies when table structure has evolved. It treats DeepChat's internal format as documentable but not stable, so you inspect first rather than hardcode assumptions.

GitHub stars: 6K
last updated: Jun 11, 2026

View on GitHub

Install

$ npx -y skills add thinkinaixyz/deepchat --skill deepchat-data-import --agent claude-code

Installs into .claude/skills of the current project.

Files

DeepChat Data Import

Overview

Use this skill to design or implement importers for DeepChat local data. Treat DeepChat's SQLite schema as an internal but documentable contract: inspect the current schema when precision matters, prefer read-only snapshots, and avoid writing to a live profile.

Workflow

Identify the source: live DeepChat profile, copied profile, sync backup, exported agent.db, or legacy chat.db.
Read references/data-locations.md to locate agent.db, sidecar files, encryption metadata, and backup paths.
Read references/sqlite-access.md before opening SQLite. Decide whether the database is unencrypted, can be unlocked through Electron safeStorage, or must ask the user for the SQLite password.
Read references/schema-reference.md for provider config, settings, session, message, and legacy table relationships.
Read references/import-recipes.md when writing extractor code, mapping DeepChat data to another app, or creating a compatibility import.