The Agent Skills Directory

[SAFE]: The skill is authored by the service provider (longbridge) and interacts exclusively with the vendor's official infrastructure through the longbridge CLI and MCP tools. No unauthorized data exfiltration or third-party connections were found.
[COMMAND_EXECUTION]: Shell commands are restricted to the primary purpose of the skill: managing and querying account information via the longbridge utility.
[DATA_EXPOSURE]: The skill handles sensitive financial information but includes explicit privacy instructions for the agent, advising it to confirm with the user before displaying exact figures if a third party might be observing. This is a security best practice for handling private data.
[CREDENTIALS_UNSAFE]: Authentication is managed through a secure login flow (longbridge auth login), preventing the need for hardcoded credentials or unsafe secret storage within the skill itself.

longbridge-positions