Give it the entry point.
Copy the starter prompt into your coding agent, or open the root skill and provide it yourself.
Open root SKILL.mdA field guide for AI builders
Your agent.
Your idea.
Avalanche.
Give your coding agent the context to build on Avalanche. Focused skills. Official sources. No guesswork about which chain you’re on.
Plain Markdown. Use with your agent of choice.
CONTEXT → CAPABILITY01—__COUNT_PAD__
agent / contextMD
# Start here. Go deep when needed.
read ./SKILL.md
└ understand the task
└ choose the right chain
└ load a focused skill
Ready to build with context.
A small context window. A much better starting point.
__COUNT__focused skills
One ecosystem.
One ecosystem.
Source-linked guidance
Verify at the source.
Verify at the source.
Agent-readable Markdown
Less setup. More building.
Less setup. More building.
THE LIBRARY / 01
The knowledge layer.
Start broad. Pull in exactly
what your next build needs.
__FILTERS__
__CARDS__
No skills match that search.
Try a broader term or reset your filters.
THE WORKFLOW / 02
Context in. Better builds out.
Load just enough.
The root skill routes your agent to the right topics: C-Chain, custom L1s, interchain messaging, and more.
Find your next skillVerify. Then build.
Follow the official sources, check live network details, and test the target network’s behavior before a production transaction.
Read the agent indexBUILT WITH A POINT OF VIEW
Avalanche-specific.
Open by design.
Not an Ethereum guide with the names swapped. This library starts with Avalanche’s own architecture, networks, and developer workflows.
An independent community project. Not affiliated with Ava Labs or the Avalanche Foundation. Source-reviewed guidance is not a security audit.
Download skills + examples