← Back to Skills Marketplace
1519
Downloads
16
Stars
2
Active Installs
36
Versions
Install in OpenClaw
/install zonein
Description
Autonomous Hyperliquid trading agent powered by smart money signals. Create, backtest, and deploy AI trading agents that track 500+ whale wallets on Hyperliq...
Usage Guidance
This skill appears internally consistent for a trading agent, but before enabling it consider: (1) Only provide a ZoneIn API key with the minimum scopes needed (prefer read-only or no-withdraw if testing). (2) Prefer HITL (human-in-the-loop) or paper-trading modes instead of fully automated 'auto' execution until you trust behavior. (3) Confirm the API base (mcp.zonein.xyz) and homepage (zonein.xyz) match official ZoneIn documentation and reputation. (4) Be aware the skill will use any ZONEIN_API_KEY found in your environment or ~/.openclaw/openclaw.json — avoid storing high-privilege keys in plaintext. (5) Review the bundled scripts (scripts/zonein.py) and sample configs (references/) yourself; rotate the key immediately if you suspect misuse. (6) If you need stronger safety, request an API key that cannot withdraw funds and test via backtests/paper trading first.
Capability Assessment
Purpose & Capability
Name/description, required binary (python3), and required env var (ZONEIN_API_KEY) match the code and SKILL.md. The script only contacts the ZoneIn API (https://mcp.zonein.xyz) and uses the API key as the primary credential. There are no unrelated env vars, binaries, or config paths requested.
Instruction Scope
SKILL.md and scripts direct the agent to call the bundled CLI (scripts/zonein.py) which makes authenticated requests to the ZoneIn API and reads ~/.openclaw/openclaw.json as an API-key fallback. Financial actions are programmatically gated behind a --confirm flag in the script (documented). Note: because the skill can execute trading commands (agent-deploy, agent-open, agent-withdraw, etc.), granting it the API key enables live trading operations — the gating helps but the agent or integrator could invoke commands with --confirm.
Install Mechanism
No installer or downloaded artifacts; this is effectively instruction + a Python script included in the skill bundle. Nothing is downloaded at install-time and no archives or third-party installers are referenced.
Credentials
Only ZONEIN_API_KEY is required, which is appropriate for a client that calls the ZoneIn API. However this single key is high-sensitivity: if it has trading/withdrawal privileges on the ZoneIn account, the skill (or anyone who can invoke it with --confirm) could execute financial actions. The skill reads the key from environment or ~/.openclaw/openclaw.json as documented.
Persistence & Privilege
always is false and the skill does not request elevated platform privileges or to modify other skills. It can be invoked autonomously (platform default), which is expected for an agent skill that can run agents — combine this with the note above about the API key when deciding to enable autonomous invocation.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install zonein - After installation, invoke the skill by name or use
/zonein - Provide required inputs per the skill's parameter spec and get structured output
Version History
v2.3.8
Update skill docs for backtest fixes
- DATA_SOURCES.md: Fix SM thresholds 60→50/55 based on real market data
- SKILL.md: Add leverage warning and minimum hold time guidance
- AGENT_CONFIG.md: Add leverage×SL caps table and precision_master SL fix
Real market data shows SM long_ratio averages 44-46% and rarely exceeds 55%.
Lower thresholds prevent false negatives while maintaining signal quality.
v2.3.7
Updated display name and description for better ClawHub search ranking. No functional changes
v2.3.6
Updated description for clarity and conciseness, emphasizing plain English agent creation, HITL trade plans via Telegram, and supported asset classes.
v2.3.5
Anti-hallucination: clarify --confirm boundary, agent-update full capabilities, withdraw sweep-only
v2.3.4
Anti-hallucination: clarify --confirm boundary, agent-update full capabilities, withdraw sweep-only
v2.3.3
docs: prioritize QR code preview over Arbiscan link in deposit format
v2.3.2
fix: replace markdown image with clickable QR link for Telegram compatibility
v2.3.1
fix: replace markdown image with clickable QR link for Telegram compatibility
v2.3.0
Zonein 2.3.0 — Major upgrade: Enhanced documentation, reference structure, command UX, and smart money analysis.
- Added extensive in-skill references: detailed COMMANDS.md, DATA_SOURCES.md, AGENT_CONFIG.md, and workflows documentation under `references/`.
- Improved SKILL.md with explicit use-cases, field descriptions, agent creation flow, and critical safety instructions.
- Updated presentation and approval logic: stricter financial command gating, clear handling guidelines for wallet and agent IDs.
- Removed legacy scripts/check_signals.py, scripts/market_overview.py, and scripts/track_trader.py.
- Now explicitly compatible with python3 and OpenClaw with ZONEIN_API_KEY.
- Rewrote Quick Reference table and when-to-use skill guides for better usability.
v2.1.11
AI now asks your trading style in plain language (no more technical type names), auto-generates all entry/exit rules from your strategy description, and always shows strategy examples. Fixed several command flag inconsistencies.
v2.1.10
zonein 2.1.10
- Expanded Quick Reference with new HIP-3 commands for DEX and asset listings, stock trading, and agent creation.
- Added support for commands: `hip3-dexs`, `hip3-assets`, and new agent-open parameters for HIP-3 stock trading.
- Documented the new `agent-pending-plans` read-only command.
- Updated financial and approval command lists to include plan rejection/approval (`agent-plan-action reject/approve --confirm`).
v2.1.9
fix trigger_conditions thresholds and exit logic for public release - SM thresholds: AND >=50, OR >=55, EXIT AND >=55 (min 50% policy)
- Exit logic: all examples changed from OR to AND (require 2+ confirmations)
- Added exit OR premature exit warning and entry/exit consistency rule
- Fixed agent-create examples with OR confirmations + AND exits
v2.1.8
No user-visible changes detected in this release.
- Version number updated to 2.1.8.
- No modifications to documentation or code from previous version.
v2.1.7
feat: add agent-update-sl-tp CLI command
Add cmd_agent_update_sl_tp handler + argparse entry
Update module docstring with new command
POST /agents/{id}/update-sl-tp endpoint
v2.1.6
remove market.taker_buy_sell_ratio from all references — field unavailable in dashboard snapshots
v2.1.5
feat: redesign agent creation flow
zonein.py: add --withdrawal-addresses to agent-create
zonein.py: fix agent-update old field names (methodology→trading_strategy, entry_strategy→custom_rules, exit_framework→risk_management)
zonein.py: add 6 missing params to agent-update (trigger-conditions, trading-risk, signal-weights, prompt-config, execution-mode, withdrawal-addresses)
SKILL.md: remove confusing A/B/C/D signal type options
SKILL.md: add 20 diverse strategy examples with specific metrics (SM + TA + Market)
references/schema.md: update to V2 agent config schema
references/strategy.md: update trading flow to V2 (trigger_conditions, composite signals, HITL)
v2.1.4
docs: improve SKILL.md Q4 trading philosophy, add Q5 execution mode, fix Q numbering
Rewrite Q4: from obvious data-source question (A/B/C) to insightful trading philosophy (Trend Following, Early Entry, Contrarian, High-Conviction, Custom)
Add Q5: execution mode with clear auto vs hitl comparison table and recommendation
Add --withdrawal-addresses and --execution-mode to agent-create params table
Fix Q numbering: Q6=additional notes, Q7=withdrawal address
v2.1.3
No code or documentation changes detected in this release.
- Version bump from 2.1.0 to 2.1.3 only; functionality and documentation remain the same.
v2.1.2
docs: add threshold ranges, withdrawal Q6, Telegram prompt to SKILL.md
- Add Recommended Threshold Ranges table for all SM/TA/Market metrics
- Update intent->condition mapping with realistic thresholds
- Add Q6: ask withdrawal_addresses before agent creation
- Add Telegram setup prompt in Step 3 after deploy
v2.1.1
feat: align CLI and SKILL.md with all MCP server capabilities
Metadata
Frequently Asked Questions
What is Hyperliquid Trading Agent?
Autonomous Hyperliquid trading agent powered by smart money signals. Create, backtest, and deploy AI trading agents that track 500+ whale wallets on Hyperliq... It is an AI Agent Skill for Claude Code / OpenClaw, with 1519 downloads so far.
How do I install Hyperliquid Trading Agent?
Run "/install zonein" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Hyperliquid Trading Agent free?
Yes, Hyperliquid Trading Agent is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Hyperliquid Trading Agent support?
Hyperliquid Trading Agent is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Hyperliquid Trading Agent?
It is built and maintained by Phu Trinh (@phutt-bwai); the current version is v2.3.8.
More Skills