Feature Request: Integration between Claude App (Projects / Deep Research) and Claude Code

Resolved 💬 2 comments Opened Feb 4, 2026 by MoonOMT Closed Mar 5, 2026

Summary

Claude App and Claude Code are both Anthropic products, but there is no integration between them. Users who run Deep Research or use Projects on claude.ai cannot automatically access those results from Claude Code. This creates a frustrating workflow gap.

Problem

  • Deep Research results on claude.ai are locked inside the browser with no way to export or access them programmatically.
  • Claude Code has powerful local file management, screening tools, and automation capabilities, but cannot access Claude App data.
  • Users must manually copy-paste results between the two, which defeats the purpose of having an AI assistant.
  • The conversation export feature exports ALL data, not per-project, making it impractical for targeted workflows.

Proposed Solution

  1. Per-Project export: Allow exporting conversations from a specific Project on claude.ai (not just full account export).
  2. API access to Projects / Deep Research: Expose an API endpoint so Claude Code (or MCP servers) can read Project conversation history.
  3. Shared context layer: Allow Claude Code to reference or pull results from claude.ai Projects, enabling seamless workflows like: research on App → analyze/filter on Claude Code → save locally.

Use Case

A user has a Claude Project template for investment Deep Research. They run analysis on multiple companies. They want to:

  • Automatically save all results as local Markdown files
  • Filter results by specific criteria (e.g., BUY recommendations)
  • Use Claude Code's local tools for further analysis

Currently, this requires manual copy-paste for each company, which is unacceptable for a seamless AI workflow.

Impact

This integration would make the Claude ecosystem significantly more powerful as a unified platform rather than disconnected tools.

View original on GitHub ↗

This issue has 2 comments on GitHub. Read the full discussion on GitHub ↗