supabase-checkpointer
Durable LangGraph checkpointing via Supabase REST API (PostgREST) — no direct Postgres connection needed. Use when deploying LangGraph graphs to ephemeral platforms (Railway, Fly, Cloud Run) where in-memory state is lost on redeploy, and you need interrupt/resume to survive process death. Works with any Supabase project using the existing service role key.
Why use this skill?
Enable persistent state for LangGraph agents on ephemeral platforms. Use Supabase REST API for durable checkpoints and process-death survival.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/brandonwadepackard-cell/supabase-checkpointerWhat This Skill Does
The supabase-checkpointer is a specialized persistence layer for LangGraph agents designed specifically for modern cloud environments. By leveraging the Supabase PostgREST API, this skill allows your LangGraph checkpoints—which store the state, history, and metadata of your AI agent's execution flow—to persist outside of the volatile memory of your server. When you deploy agents to ephemeral environments like Railway, Fly.io, or Google Cloud Run, processes are frequently restarted or destroyed, which would normally wipe your agent's memory. This skill acts as a durable storage bridge, ensuring that your AI can pause, interrupt, and resume long-running workflows across process deaths without requiring a direct, high-latency database connection string.
Installation
To integrate this into your workflow, execute the following command in your terminal:
clawhub install openclaw/skills/skills/brandonwadepackard-cell/supabase-checkpointer
Before using the skill, ensure you have executed the required SQL schema inside your Supabase project's SQL editor. The skill relies on two primary tables: langgraph_checkpoints for state storage and langgraph_writes for tracking task execution status. Once the tables are created, instantiate the checkpointer in your LangGraph pipeline using your Supabase client, as shown in the developer documentation.
Use Cases
This skill is ideal for any production-grade agentic application. It is primarily used when building complex multi-step workflows that must survive redeployments. For example, if your agent is managing a long-running customer ticket that requires human-in-the-loop approval, the state must be saved to ensure the interaction survives a server reboot. It is also perfect for distributed agent systems where multiple worker nodes need to pick up tasks from a centralized state, ensuring thread safety and data persistence across a fleet of stateless containers.
Example Prompts
- "Check the current checkpoint state for thread 'customer-support-123' using the Supabase checkpointer and tell me what step the agent is currently stuck on."
- "Set up my LangGraph builder to use the SupabaseCheckpointer as the primary persistence layer instead of the default MemorySaver."
- "List all recent checkpoint versions for my active task thread to help me identify where the process failed during the last deployment cycle."
Tips & Limitations
To maximize reliability, implement the recommended fallback chain. Start with Supabase for durability, fallback to a local SQLite database for intermediate restarts, and finally to memory-only storage as a last resort. Keep in mind that while this avoids direct Postgres connections, you are subject to standard Supabase API latency; ensure your network configuration allows communication with your specific Supabase region. Always use the service role key carefully, as it grants elevated access to your data tables.
Metadata
Not sure this is the right skill?
Describe what you want to build — we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-brandonwadepackard-cell-supabase-checkpointer": {
"enabled": true,
"auto_update": true
}
}
}Tags(AI)
Flags: external-api, data-collection
Related Skills
domain-email-forwarding
Set up email forwarding for custom domains to receive verification codes, password resets, and other emails at a domain you own but don't actively use for email. Covers GoDaddy, Namecheap, Cloudflare Email Routing, and ImprovMX. Use when recovering accounts linked to inactive domain emails, setting up catch-all forwarding, or routing domain email to Gmail/Outlook without paying for email hosting.
platform-api-connector
Connect to social media and content platform APIs by navigating developer portals, creating apps, obtaining OAuth tokens, and storing credentials. Covers Facebook Graph API, Instagram Business API, YouTube Data API, Twitter/X API v2, and TikTok Content Posting API. Use when setting up API access for any social platform, refreshing expired OAuth tokens, or debugging authentication flows.
hume-evi-langgraph
Integrate Hume EVI voice AI with LangGraph using interrupt/resume patterns. Use when building voice-based AI agents that need Twilio call handling, Hume EVI persona creation, transcript fetching with emotion extraction, and LangGraph state management across the call lifecycle. Covers dynamic Hume config creation, TwiML generation, webhook handling, chat_group event fetching, and emotion timeline extraction.
ai-persona-engine
Build emotionally intelligent AI personas for voice and chat roleplay using actor-direction prompts instead of technical specifications. Use when creating AI characters that need to sound like real people, not chatbots — covers the Doctrine of Embodied Intelligence, brevity-first voice design, 5-layer conversation intelligence, elemental archetypes, self-auditing persona quality, and the principle that resistance equals fewer words.
social-media-platform
Build a plugin-based social media management platform with multi-platform publishing, content calendar, brand voices, AI content generation via LangGraph, and analytics. Use when building social media tools, content scheduling systems, or multi-platform publishing pipelines. Covers Facebook, Instagram, YouTube, Twitter/X, TikTok integration patterns.