Claude Code Plugins

Community-maintained marketplace

Feedback

Incrementally import soul data from SQLite or shared chitta files. Adds to existing data.

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name migrate
description Incrementally import soul data from SQLite or shared chitta files. Adds to existing data.
execution task

Soul Migration (Incremental)

Import soul data incrementally - new nodes are added to existing data.

Sources

  1. SQLite soul.db - Legacy cc-soul format (chitta_migrate)
  2. Shared chitta files - From another user (chitta_import)

Execute

Task(
  subagent_type="general-purpose",
  description="Import soul data incrementally",
  prompt="""
Import soul data incrementally. Ask user which source to import from.

## 1. Determine Source

Ask user: "Import from (1) SQLite soul.db or (2) shared chitta files?"

## 2a. SQLite soul.db Import

Find binaries:
```bash
PLUGIN_DIR=$(ls -d ~/.claude/plugins/cache/genomewalker-cc-soul/soul/*/chitta 2>/dev/null | tail -1)
MIGRATE_BIN="$PLUGIN_DIR/build/chitta_migrate"
MODELS_PATH="$PLUGIN_DIR/models"

# Build if missing
if [ ! -f "$MIGRATE_BIN" ]; then
  cd "$PLUGIN_DIR/build" && make chitta_migrate
fi

Dry run:

$MIGRATE_BIN --dry-run --verbose \
  --soul-db ~/.claude/mind/soul.db \
  --output ~/.claude/mind/chitta \
  --model "$MODELS_PATH/model.onnx" \
  --vocab "$MODELS_PATH/vocab.txt"

If approved, run import:

$MIGRATE_BIN --verbose \
  --soul-db ~/.claude/mind/soul.db \
  --output ~/.claude/mind/chitta \
  --model "$MODELS_PATH/model.onnx" \
  --vocab "$MODELS_PATH/vocab.txt"

2b. Chitta Files Import

Ask user for source path (base path without .hot/.cold suffix).

Find binaries:

PLUGIN_DIR=$(ls -d ~/.claude/plugins/cache/genomewalker-cc-soul/soul/*/chitta 2>/dev/null | tail -1)
IMPORT_BIN="$PLUGIN_DIR/build/chitta_import"
MODELS_PATH="$PLUGIN_DIR/models"

# Build if missing
if [ ! -f "$IMPORT_BIN" ]; then
  cd "$PLUGIN_DIR/build" && make chitta_import
fi

Dry run:

$IMPORT_BIN --dry-run --verbose \
  --source /path/to/shared/chitta \
  --target ~/.claude/mind/chitta \
  --model "$MODELS_PATH/model.onnx" \
  --vocab "$MODELS_PATH/vocab.txt"

If approved, run import:

$IMPORT_BIN --verbose \
  --source /path/to/shared/chitta \
  --target ~/.claude/mind/chitta \
  --model "$MODELS_PATH/model.onnx" \
  --vocab "$MODELS_PATH/vocab.txt"

3. Verify

After import:

  • mcp__plugin_cc-soul_cc-soul__soul_context(format="json") - Check node counts
  • mcp__plugin_cc-soul_cc-soul__recall(query="wisdom", limit=3) - Verify search

4. Report

  • Source type and path
  • Nodes imported by type
  • Current soul state """ )