snek.share

Share Conversation with Snek

Export your current Claude conversation to snek's session system so snek can understand the full context of your work.

---

STEP 1: Find the Active Session

First, read .snek/active.json to find the current active session:

cat .snek/active.json

This file contains:

{
  "schema": 1,
  "id": "session-uuid-here",
  "path": "sessions/session-uuid-here"
}

The session directory is .snek/sessions/{id}/ where {id} is the UUID from the id field.

Example: If id is "b4c12a8f-5a92-4e92-8d3c-7f9a1c2e5b8a", the session directory is:

.snek/sessions/b4c12a8f-5a92-4e92-8d3c-7f9a1c2e5b8a/

---

STEP 2: Navigate to Session Context Folder

The session directory structure is:

.snek/sessions/{session-id}/
├── session.json           # Session metadata (don't edit)
├── code_snippets.json     # Code file references (you'll edit this)
└── context/               # Markdown context files (you'll add files here)
    ├── architecture.md
    ├── conventions.md
    └── ... (your new files)

Your target locations:

• Markdown context: .snek/sessions/{session-id}/context/
• Code references: .snek/sessions/{session-id}/code_snippets.json

---

STEP 3: Create/Update Markdown Context Files

Location: .snek/sessions/{session-id}/context/

3A. Check Existing Files

List all existing markdown files:

ls .snek/sessions/{session-id}/context/

Read each file to see what context is already documented. DO NOT duplicate information that's already there.

3B. Create New Context File(s)

Create a new markdown file with a descriptive name that reflects this conversation. Good names:

• claude_conversation_2025_11_07.md
• lsp_implementation_discussion.md
• session_management_context.md

File naming rules:

• Use lowercase with underscores
• Be descriptive and specific
• Files are read alphabetically, so prefix with numbers if order matters (e.g., 01_overview.md, 02_details.md)