← Back to Skills Marketplace
risk art agent
by
riskiagung-prog
· GitHub ↗
· v0.1.0
· MIT-0
240
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install riskagent
Description
AI-powered crypto trading agent and LLM gateway via natural language. Use when the user wants to trade crypto, check portfolio balances, view token prices, t...
Usage Guidance
This skill appears to do what it says (a Bankr trading + LLM gateway), but exercise caution before installing or running it. Key points:
- The SKILL.md will ask you to generate/provide API keys (bk_... and optional BANKR_LLM_KEY) and may write them into ~/.bankr/config.json and into other tool configs (notably ~/.openclaw/openclaw.json). Only supply keys you control and prefer creating a least-privilege key (read-only) unless you explicitly want trading and transfer capabilities.
- The agent can submit arbitrary transactions and enable read-write permissions; do not enable --read-write unless you fully trust the agent and understand the financial risk. Prefer read-only for research/balances.
- The skill's registry metadata did not list required env vars, despite the instructions referencing several; treat that omission as a red flag and verify manually which env vars/configs the CLI will write or require.
- Carefully review any Terms of Service before using --accept-terms on the user's behalf. Never share OTPs or API keys unless you intend the agent to act with those credentials.
- If you want to limit risk: create a dedicated Bankr account/API key with minimal permissions and small balances, avoid automated --install steps that write to other agent configs, and test with tiny amounts or on testnets first.
If you want, I can: summarize exactly which commands the SKILL.md will run for a given flow (login, llm setup, token transfer), or produce a checklist of consent prompts the agent should present before performing each high-risk action.
Capability Analysis
Type: OpenClaw Skill
Name: riskagent
Version: 0.1.0
The 'bankr' skill bundle provides a comprehensive suite for crypto trading and DeFi operations, including high-risk capabilities such as executing arbitrary blockchain transactions via raw hex calldata (documented in 'references/arbitrary-transaction.md' and 'references/sign-submit-api.md'). While these features are aligned with the stated purpose of a trading agent, they represent a significant attack surface for prompt injection that could lead to unauthorized fund transfers. The skill also manages sensitive API keys and relies on an external CLI (@bankr/cli) and remote API (api.bankr.bot), introducing substantial third-party risk and potential for financial loss if the agent is misdirected.
Capability Assessment
Purpose & Capability
The skill's name/description (crypto trading + LLM gateway) matches the content: it relies on the Bankr CLI/REST API and exposes trading, wallet, and LLM features. Requiring the 'bankr' binary is expected. However, the SKILL metadata declares no required env vars while the instructions clearly expect/advise setting BANKR_API_KEY, BANKR_LLM_KEY, and other env vars/config entries — this is an inconsistency between declared requirements and what the skill actually uses.
Instruction Scope
The SKILL.md instructs the agent to perform sensitive flows: headless login (asking the user for OTP), generate API keys (read-only vs read-write), sign/submit arbitrary raw transactions, and enable LLM gateway access. It also tells the agent to run bankr llm setup openclaw --install which writes provider config (including apiKey) into ~/.openclaw/openclaw.json and to write to ~/.bankr/config.json. These actions go beyond simple querying: they store credentials and modify other tool configs and enable write-enabled access to wallets and on-chain transactions. While functionally consistent with a trading agent, the instructions allow high-risk operations (arbitrary calldata submission, leverage trading, transfers) and modify external configs — require explicit, careful user consent.
Install Mechanism
This is an instruction-only skill (no install spec), so nothing is written by the registry itself. The SKILL.md recommends installing @bankr/cli via bun/npm — a standard package install approach. No arbitrary download URLs or extract operations are present in the manifest.
Credentials
The registry metadata declares no required environment variables, but the instructions repeatedly reference and instruct users to set BANKR_API_KEY, BANKR_LLM_KEY, ANTHROPIC_* env vars and to store API keys in config files. Requesting API keys and LLM keys is expected for a trading/LLM gateway, but the omission in declared requirements is a mismatch. Additionally, instructions suggest storing keys in multiple locations (~/.bankr/config.json and writing into ~/.openclaw/openclaw.json), increasing the attack surface and risk of accidental credential exposure.
Persistence & Privilege
The skill does not set always:true, but its instructions explicitly direct modifying other agent/tool configuration files (e.g., writing provider/apiKey into ~/.openclaw/openclaw.json) and exporting credentials to shell profiles. Modifying another agent's config and persisting API keys in third-party config files is a meaningful privilege escalation relative to an ephemeral instruction-only skill and should be performed only with explicit user consent. The skill also enables operations (read-write agent API, sign & submit raw transactions, leverage) that carry financial risk.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install riskagent - After installation, invoke the skill by name or use
/riskagent - Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.1.0
Initial release of the Bankr skill for AI-powered crypto trading and DeFi operations:
- Enables crypto trading, portfolio checks, token price queries, transfers, NFT management, leverage trading, Polymarket bets, token deployments, and automated trading via natural language.
- Supports interaction through the Bankr CLI tool (installation and guided login flows provided) or direct REST API calls.
- Handles user onboarding, API key creation (including headless flows and user preference prompts), and seamless EVM/Solana wallet provisioning.
- Added capability to access LLM models through the Bankr LLM gateway, funded by the user's Bankr wallet.
- Compatible with Base, Ethereum, Polygon, Solana, and Unichain networks.
Metadata
Frequently Asked Questions
What is risk art agent?
AI-powered crypto trading agent and LLM gateway via natural language. Use when the user wants to trade crypto, check portfolio balances, view token prices, t... It is an AI Agent Skill for Claude Code / OpenClaw, with 240 downloads so far.
How do I install risk art agent?
Run "/install riskagent" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is risk art agent free?
Yes, risk art agent is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does risk art agent support?
risk art agent is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created risk art agent?
It is built and maintained by riskiagung-prog (@riskiagung-prog); the current version is v0.1.0.
More Skills