flapskill
Audited by Socket on Mar 12, 2026
3 alerts found:
Obfuscated Filex2AnomalyThe script itself is not malicious in intent or behavior, but it poses a significant operational security risk because it generates private keys and stores them in plaintext in a predictable file in the current working directory while also printing addresses and the file path to stdout/stderr. Treat outputs as highly sensitive. Recommendations: do not run in shared/CI environments; avoid printing secrets; store private keys in secure keystores or encrypted files; set restrictive file permissions (e.g., 600) and secure deletion if temporary; consider prompting before writing, and avoid predictable filenames or leaking paths to logs.
This is a fund-collection/aggregation tool that consolidates both ERC20 tokens and remaining BNB from multiple worker accounts into a single target address. While legitimate use exists in admin consolidation scenarios, the script introduces significant security risk: it automates sweeping funds from multiple private-key-controlled wallets without explicit per-wallet authorization prompts or safeguards. The primary risk stems from handling plaintext private keys in a local file and executing automated transfers, which could be abused if keys are compromised or misused. Strong access controls, key management, and optional dry-run/approval workflows are strongly recommended before use in any production environment.
The skill exhibits strong purpose-capability alignment for token creation and automated market-making, but its footprint markedly increases security risk due to autonomous on-chain actions, reliance on private keys, unlimited USDT approvals, and generation/management of worker keys. The data flows enable substantial credential access and funds movement with minimal per-action user prompts. Overall classification: Suspicious-to-High risk (leaning towards high risk) due to credential exposure, autonomous asset movement, and multiple potential supply-chain/remote-script risks. Benign only if strictly audited, with hardened controls and explicit per-action user approvals.