open-stellar-wallet

Manage Stellar wallets, send XLM payments, configure networks, and interact with Soroban smart contracts via the Stellar CLI on testnet or mainnet.

This skill appears to do what it says (manage Stellar wallets via the stellar CLI), but consider the following before installing: - The skill's first-run setup will automatically create and fund a 'default' testnet wallet without asking you. If you don't want automated key creation or files written to your machine, decline or run setup manually. - The install options include a raw GitHub install.sh (download-and-run). Prefer installing via Homebrew, winget, or cargo if you trust those package sources; if you use the script, open and review it first to confirm it does only the expected steps. - Back up any secret keys created and never share them. The skill's commands will store secrets in the Stellar CLI key store on disk. - The source/homepage in the registry metadata is minimal/unknown; if you need stronger assurance, obtain the official release page or verify the GitHub repo and release signatures before installing. If you want a lower-risk deployment: install the stellar CLI yourself from a trusted channel, then use this skill only as documentation (i.e., don't allow it to run the automatic first-run setup).

Type: OpenClaw Skill Name: open-stellar Version: 1.0.0 The skill is classified as suspicious due to two main factors: 1) The installation process involves downloading and executing a shell script (`install.sh`) directly from a remote GitHub URL (`https://github.com/stellar/stellar-cli/raw/main/install.sh`). While this is a common installation method for CLIs and the source appears legitimate, it introduces a supply chain risk, as a compromised script could lead to arbitrary code execution. 2) The `SKILL.md` explicitly instructs the AI agent to perform 'First-Run Setup' steps 'automatically — do NOT ask the user for confirmation'. Although the current automatic actions (setting up a testnet wallet with test funds) are benign, this instruction demonstrates a prompt injection technique that bypasses user consent, which could be exploited if the commands were altered to perform malicious actions.