Universal Trading

Execute cross-chain token trading on EVM and Solana with Particle Network Universal Account SDK. Use when users ask to set up universal-account-example, buy...

This skill will generate or import a wallet and create a .env containing PRIVATE_KEY and Particle project credentials, patch example code, run npm installs, and call Particle Network APIs (including an automatic invite-bind flow using invite code '666666' and embedded demo PROJECT_* keys). Before installing or running it: 1) do not run the auto-init in a directory with important files; run it in an isolated test workspace; 2) inspect scripts/init.sh and setup-wizard.sh and consider running them step-by-step rather than auto-run; 3) replace the hardcoded PROJECT_* demo credentials with your own Particle project credentials if you plan to use real funds; 4) note that .env will contain your private key in plaintext (scripts try to chmod 600) — consider using a safer key management workflow (hardware wallet, ephemeral test key, or encrypting the key); 5) disable auto-invite bind by setting DISABLE_AUTO_INVITE_BIND=1 if you don't want the skill to register/bind your wallet with the external invite endpoint; and 6) test with minimal amounts only after verifying network endpoints and code. The main incoherence is that the registry declares no required env vars while the runtime requires sensitive credentials — treat that as a red flag and proceed cautiously.

Type: OpenClaw Skill Name: universal-trading Version: 0.1.0 This skill is highly suspicious due to its direct handling of cryptographic private keys and execution of financial transactions. The `SKILL.md` and `agents/openai.yaml` instruct the AI agent to generate or import private keys, store them in `universal-account-example/.env`, and use them to sign blockchain transactions and interact with external APIs (e.g., `particle.network`). The `scripts/init.sh` and `scripts/setup-wizard.sh` are particularly risky as they accept private keys directly as command-line arguments, which is a significant vulnerability for shell injection if the agent's input is not perfectly sanitized. Additionally, `scripts/bootstrap-example.sh` introduces a supply chain risk by cloning and executing code from a remote GitHub repository. While the skill's explicit purpose is to enable trading, the inherent risks and potential for misuse or exploitation of these capabilities warrant a 'suspicious' classification.