| name | help |
| description | Shows all available eth-agents skills with trigger phrases and usage examples. |
| triggers | ["help","what can you do","list skills","how do I","show me the skills","what skills are available","eth-agents help"] |
| agent | orchestrator |
| model | haiku |
Purpose
Returns a formatted guide of all available skills with their trigger phrases and example invocations. No agents are spawned — this is a direct orchestrator response.
Workflow
- Orchestrator (haiku) returns the help text directly
- No agents spawned, no MCP calls needed
Usage
"help"
"what can you do?"
"show me all skills"
"how do I audit a contract?"
# Output (inline, no files):
Formatted table of all skills with triggers and examples
Output Format
# eth-agents — Available Skills
| Skill | What it does | Example |
|---|---|---|
| audit | Security analysis | "audit this contract" |
| tdd | Test-driven development | "write tests first for an ERC20" |
| build-contract | Write → test → audit → optimize | "build a staking contract" |
| deploy | Multichain deploy + verify | "deploy to Arbitrum" |
| gas-optimize | Reduce gas costs | "optimize gas for the Vault" |
| fork-test | Mainnet fork tests | "test against mainnet" |
| full-protocol | End-to-end build | "build the full protocol" |
| learn | Save a reusable pattern | "remember this pattern" |
| help | This guide | "help" |
## Quick Start
- Describe what you want in plain English — no commands to memorize
- eth-agents auto-detects Foundry vs Hardhat, your frontend framework, and your package manager
- All agents follow checks-effects-interactions and never hardcode addresses
Configuration
No configuration. Static response only.