Add simple working chat that bypasses Marimo AI panel

Add zstd dependency for Ollama install

Start Ollama and configure Marimo AI

Add Ollama config proxying to ZeroClaw

Install Ollama, remove Pi/pi-acp

Use mo.ui.chat like official marimo chatbot example

Copy ZeroClaw AI config on startup

Configure ZeroClaw as Marimo AI provider

New app.py with ZeroClaw chat + Pi RPC placeholder

Name pi-acp as opencode so Marimo recognizes it

Copy marimo config on startup

Add pre-configured marimo.toml with Pi provider

Force reinstall pydantic-ai with version pin

Start pi-acp via stdio-to-ws bridge on port 3017

Add stdio-to-ws bridge for Marimo ACP

Add pydantic-ai for Marimo compatibility

Add Pi coding agent install (required by pi-acp)

Initialize Pi before starting pi-acp

Add Pi config setup - use ZeroClaw as provider

Add Pi config to use ZeroClaw as provider

Start pi-acp ACP server for Marimo AI

Add pi-acp install for ACP support

Switch to marimo edit mode with --no-token

Remove .zeroclaw copy - using env var instead

Use ZEROCLAW_GATEWAY_PORT env var to set port 42617

Add ZeroClaw config copy to set port 42617

Add ZeroClaw config to set port to 42617

Remove --port flag from ZeroClaw (not supported)

Fix: ZeroClaw port 7862, launch Marimo not Gradio

Remove mom-infra build - not needed, speeds up build time

Update app.py with clean VerdantClaw UI

v2.8: FORCE FULL REBUILD - clear cache

v2.3: Full deployment with app.py

v2.2: Remove invalid --no-watch flag

v2.5.4: Restore Marimo UI on 7860 + Pi Agent persistence fix

v2.5.3: Restore container persistence with background tail wait + UI port fix

v2.5.2: Force Pi Agent into RPC Mode

v2.5.1: Fix npm pi lifecycle script

v2.5.0: Memgraph Path Fix + API Key Framework

v2.4.0: Secure Deployment - Zero Bloat

initial commit