lp-agent

SUSPICIOUS: the skill is mostly purpose-aligned and uses largely official same-org infrastructure, but it enables autonomous on-chain trading, asks users to paste private keys into local scripts, uses weak/default local credentials in examples, and depends on unreviewed security-sensitive scripts plus mutable development images. No clear evidence of credential exfiltration to third-party endpoints was shown, so this is high-risk workflow automation rather than confirmed malware.

No clear malicious or covert supply-chain behavior is evident. The code is intended to generate a local LP executor dashboard, but it handles API credentials insecurely when using HTTP or an untrusted HUMMINGBOT_API_URL, defaults to admin/admin, and may enable stored XSS through unescaped API data in generated HTML/JavaScript. The supplied fragment is also syntactically incomplete and would require correction before execution.

The code appears to be a legitimate Hummingbot API management CLI and contains no clear malware or supply-chain backdoor behavior. It does handle credentials and can execute trading actions. Using a configurable unencrypted HTTP URL and default admin/admin credentials creates a meaningful security risk, especially if the environment or .env files are tampered with or the API is remote. Restrict API URL configuration, require strong credentials, and use HTTPS for non-local communication.

No clear malicious behavior or supply-chain payload is present. The code is a readable administrative CLI intended to control a Hummingbot API. It does handle credentials and can perform financially consequential actions. Plain HTTP and default admin/admin credentials create a meaningful security weakness, while unencoded user-controlled URL components create request-construction risks. Use authenticated HTTPS, non-default credentials, strict API access controls, and URL encoding/validation for names and query parameters.

No clear malware, data theft, persistence, or reverse-shell behavior is present in the supplied script. The main supply-chain risk is that it clones and later pulls mutable remote code, then executes its Makefile targets without pinning or verification. The default credentials are insecure, and the root-created sudo shim plus configurable recursive deletion warrant review and hardening before production use.

The script is a legitimate Gateway setup utility with no clear embedded malware or covert exfiltration behavior. It has meaningful security risks: sourcing .env files executes arbitrary shell code, default admin/admin credentials are weak, HTTP can expose credentials, and arbitrary Docker images and RPC endpoints can be supplied. Use only trusted configuration and images, require authenticated HTTPS, validate inputs, and avoid dev_mode in production.

The code appears to be a legitimate wallet-management API client rather than malware. It intentionally handles highly sensitive API credentials and blockchain private keys and sends them to the configured HUMMINGBOT_API_URL. Security risk is elevated if that URL or local environment files can be modified, and HTTPS should be required for non-local connections. No clear malicious behavior or obfuscation is present.

The code is a legitimate service-health check and contains no clear intentional malware in the shown fragment. However, sourcing an arbitrary .env file is a significant command-execution risk if an attacker can modify any searched file. Default admin credentials and the HTTP default can also expose authentication material. Use a safe configuration parser, secure non-default credentials, HTTPS where applicable, and validate the configured endpoint. The behavior of check_api.sh remains unreviewed.