neondb-skillManage Neon serverless Postgres databases. Create projects, branches, databases, and execute queries. Perfect for agent workflows needing persistent storage with branching (like git for databases), scale-to-zero, and instant provisioning.
Install via ClawdBot CLI:
clawdbot install clawbot-ved/neondb-skillNeon is serverless Postgres — scales to zero, branches like git, instant provisioning. Perfect for AI agents needing databases without ops overhead.
# Homebrew (recommended)
brew install neonctl
# Or npm
npm i -g neonctl# Interactive (opens browser)
neonctl auth
# Or with API key (get from console.neon.tech)
export NEON_API_KEY=your_api_key_hereneonctl projects create --name "my-agent-db"# List all projects
neonctl projects list
# Create project
neonctl projects create --name "project-name"
# Delete project
neonctl projects delete <project-id>
# Get project details
neonctl projects get <project-id># List branches
neonctl branches list --project-id <project-id>
# Create branch (fork from main)
neonctl branches create --project-id <project-id> --name "dev-branch"
# Create branch from specific point
neonctl branches create --project-id <project-id> --name "restore-test" --parent main --timestamp "2024-01-15T10:00:00Z"
# Reset branch to parent
neonctl branches reset <branch-id> --project-id <project-id> --parent
# Delete branch
neonctl branches delete <branch-id> --project-id <project-id>
# Compare schemas
neonctl branches schema-diff --project-id <project-id> --base-branch main --compare-branch dev# List databases
neonctl databases list --project-id <project-id> --branch <branch-name>
# Create database
neonctl databases create --project-id <project-id> --branch <branch-name> --name "mydb"
# Delete database
neonctl databases delete <db-name> --project-id <project-id> --branch <branch-name># Get connection string (default branch)
neonctl connection-string --project-id <project-id>
# Get connection string for specific branch
neonctl connection-string <branch-name> --project-id <project-id>
# Pooled connection (recommended for serverless)
neonctl connection-string --project-id <project-id> --pooled
# Extended format (with all details)
neonctl connection-string --project-id <project-id> --extended# List roles
neonctl roles list --project-id <project-id> --branch <branch-name>
# Create role
neonctl roles create --project-id <project-id> --branch <branch-name> --name "app_user"# Get connection string and connect
neonctl connection-string --project-id <project-id> | xargs psql
# Or direct
psql "$(neonctl connection-string --project-id <project-id>)"# Get the string
CONNECTION_STRING=$(neonctl connection-string --project-id <project-id> --pooled)
# Use in any Postgres client
psql "$CONNECTION_STRING" -c "SELECT * FROM users LIMIT 5;"Set context to avoid passing --project-id every time:
# Set project context
neonctl set-context --project-id <project-id>
# Now commands use that project automatically
neonctl branches list
neonctl databases list
neonctl connection-string# Create project for org
neonctl projects create --name "website-org-db" -o json
# Create production branch (main is created by default)
# Create dev branch for testing
neonctl branches create --name "dev" --project-id <id>
# Get connection strings
neonctl connection-string main --project-id <id> --pooled # for prod
neonctl connection-string dev --project-id <id> --pooled # for dev# Connect and create schema
psql "$(neonctl cs --project-id <id>)" <<EOF
CREATE TABLE leads (
id SERIAL PRIMARY KEY,
business_name VARCHAR(255) NOT NULL,
category VARCHAR(100),
location VARCHAR(255),
phone VARCHAR(50),
email VARCHAR(255),
website VARCHAR(255),
status VARCHAR(50) DEFAULT 'identified',
priority VARCHAR(20) DEFAULT 'medium',
notes TEXT,
created_at TIMESTAMP DEFAULT NOW(),
updated_at TIMESTAMP DEFAULT NOW()
);
CREATE INDEX idx_leads_status ON leads(status);
CREATE INDEX idx_leads_category ON leads(category);
EOF# Create a branch to test schema changes
neonctl branches create --name "schema-experiment" --project-id <id>
# Test your changes on the branch
psql "$(neonctl cs schema-experiment --project-id <id>)" -c "ALTER TABLE leads ADD COLUMN score INT;"
# If it works, apply to main. If not, just delete the branch
neonctl branches delete schema-experiment --project-id <id># JSON (for parsing)
neonctl projects list -o json
# YAML
neonctl projects list -o yaml
# Table (default, human-readable)
neonctl projects list -o table# API key (required if not using `neonctl auth`)
export NEON_API_KEY=your_key
# Default project (alternative to set-context)
export NEON_PROJECT_ID=your_project_idneonctl me -o json 2>/dev/null && echo "Authenticated" || echo "Need to run: neonctl auth"# One-liner query
psql "$(neonctl cs)" -c "SELECT COUNT(*) FROM leads WHERE status='contacted';"psql "$(neonctl cs)" -c "COPY (SELECT * FROM leads) TO STDOUT WITH CSV HEADER" > leads.csvpsql "$(neonctl cs)" -c "\COPY leads(business_name,category,location) FROM 'import.csv' WITH CSV HEADER"--pooled connection string for serverless workloadsneonctl meneonctl authGenerated Mar 1, 2026
An AI agent collects business leads from web scraping and social media, storing them in a NeonDB database. The agent creates separate branches for validated leads versus raw data, allowing safe experimentation with data enrichment algorithms without affecting production data. The scale-to-zero feature ensures no costs during idle periods between lead collection cycles.
A SaaS platform uses NeonDB to provide isolated database branches for each customer's data within a single project. When onboarding new clients, the agent instantly provisions a new branch with pre-configured schemas. The branching capability allows safe testing of feature updates on customer-specific branches before rolling out to all users via the main branch.
An AI agent manages inventory levels and processes orders by interacting with a NeonDB database. It creates experimental branches to test new pricing algorithms or inventory optimization strategies without affecting live operations. The connection pooling handles sudden traffic spikes during flash sales while scale-to-zero reduces costs during off-peak hours.
A healthcare AI agent manages anonymized patient records in NeonDB with strict data isolation using branch-based access controls. The agent creates audit branches to track all data modifications and experimental branches for research analysis without compromising production patient data. Instant provisioning allows rapid setup for new clinical studies.
An AI financial advisor agent stores client portfolio data in NeonDB, using branches to create personalized investment scenarios for each client. The agent can test different risk models on branch copies before recommending strategies, with schema-diff capabilities to track changes between strategy versions. The generous free tier accommodates initial testing phases.
Offer managed NeonDB instances specifically configured for AI agent workflows, providing pre-built branching strategies and automated backup systems. Charge subscription fees based on storage usage and compute hours, with premium support for complex branching operations. The scale-to-zero feature allows competitive pricing for intermittent agent usage patterns.
Build a platform where developers can create AI agents with built-in NeonDB integration for persistent storage. Monetize through platform fees, taking a percentage of agent transaction value while providing the database infrastructure. Offer templates for common agent-database patterns like lead management or inventory tracking with optimized branching workflows.
Provide consulting services to help enterprises implement NeonDB-powered AI agents for specific business processes. Design custom branching strategies, optimize query performance, and create agent-database interaction patterns. Charge project-based fees for implementation and ongoing retainer fees for maintenance and optimization of the agent-database ecosystem.
💬 Integration Tip
Start by setting project context to avoid repeating --project-id flags, and always use pooled connection strings for serverless agent deployments to handle connection spikes efficiently during agent execution bursts.
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...
Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.
Search and analyze your own session logs (older/parent conversations) using jq.
Typed knowledge graph for structured agent memory and composable skills. Use when creating/querying entities (Person, Project, Task, Event, Document), linking related objects, enforcing constraints, planning multi-step actions as graph transformations, or when skills need to share state. Trigger on "remember", "what do I know about", "link X to Y", "show dependencies", entity CRUD, or cross-skill data access.
Ultimate AI agent memory system for Cursor, Claude, ChatGPT & Copilot. WAL protocol + vector search + git-notes + cloud backup. Never lose context again. Vibe-coding ready.
Headless browser automation CLI optimized for AI agents with accessibility tree snapshots and ref-based element selection