Overview
Blockchain Skills is a reference package that gives AI coding agents accurate knowledge of Quicknode's product suite. It provides correct endpoints, parameters, supported networks, and current best practices so that agents produce working code on the first attempt instead of guessing at API shapes or using outdated method names.
Skills uses the open skills framework by Vercel. Once installed, the skill is automatically loaded into the agent's context when working on blockchain-related tasks.
Installation
# Install the Quicknode blockchain skill for your AI coding agent
npx skills add https://github.com/quiknode-labs/blockchain-skills --skill quicknode-skill
This installs the quicknode-skill to your agent's skills directory. The installer prompts you to choose between project-level and global installation, and lets you select which AI agents to configure. If you add multiple skills for the same agent, the installer creates symlinks instead of duplicating files.
# Example prompt to the agent after installation
"Using Quicknode skills, create a script that backfills ERC-20 transfers on the Ethereum mainnet to a PostgreSQL database with Streams."
# ⏺ Skill(quicknode-skill)
# ⎿ Successfully loaded skill
# ⏺ Let me read the Streams reference for detailed filter function and PostgreSQL destination documentation.
Compatible Agents
The skill works with any agent that supports the skills framework, including:
- Claude Code
- Cursor
- Codex
- OpenCode
- Windsurf
What the Skill Covers
The quicknode-skill package covers Quicknode's core products and APIs in a single, compact reference. This includes:
The skill also includes common patterns (multi-chain setup, real-time monitoring, NFT indexing) and best practices for RPC, Streams, security, and gRPC connections.
How It Works
When an AI coding agent loads the skill, it receives structured context about Quicknode's infrastructure. The skill file includes:
- Intake questions the agent asks before generating code (which chain, read-only vs. write, RPC vs. streaming, endpoint URL)
- Safety defaults like preferring testnet when no network is specified and never accepting private keys
- Confirm-before-write rules that require explicit user confirmation before creating or modifying Streams, Webhooks, or IPFS uploads
- Code samples for each product with correct method names, parameters, and import paths
- Documentation links pointing to the right Quicknode docs page for each product
This means the agent knows to ask the right clarifying questions upfront, uses the correct API surface, and links to relevant documentation when needed.
The skill loads automatically when the agent starts a session. You can also reference it explicitly in your prompt (e.g., "use the quicknode-skill") to ensure the agent applies it for a specific task.
Additional Skills
Quicknode also maintains a Solana and Anchor skill focused specifically on Solana program development with Anchor. Install it alongside the main skill for Solana-focused work:
npx skills add https://github.com/quiknode-labs/solana-anchor-claude-skill
Resources
- Source code (open source, contributions welcome)
If you want AI tools to discover and reference Quicknode's full documentation (not just the skill's built-in knowledge), see LLM-Optimized Docs for llms.txt indexes across the platform. For managing Quicknode infrastructure through natural language, see Quicknode MCP.
We ❤️ Feedback!
If you have any feedback or questions about this documentation, let us know. We'd love to hear from you!