The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill configuration and setup.sh script define downloads of the gate-cli binary from the official Gate.com GitHub repository (github.com/gate/gate-cli). These are vendor-owned resources used for the skill's primary setup.
[PRIVILEGE_ESCALATION]: The setup.sh script attempts to acquire elevated permissions by using sudo install to move the gate-cli binary to /usr/local/bin if the local user directory installation fails.
[COMMAND_EXECUTION]: The skill relies on executing gate-cli commands for both querying data and performing trading operations. The instructions mandate a pre-execution check using the --help flag for each command to ensure all required parameters are provided before the actual execution.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes external data from Gate TradFi market and account APIs, creating a potential surface for indirect injection if malicious instructions are embedded in asset names or market descriptions.
Ingestion points: External data is ingested via the output of gate-cli commands in all query sub-modules (query-orders.md, query-positions.md, query-market.md, and query-assets.md).
Boundary markers: Absent. The workflow instructions do not define protective delimiters or warnings for the agent to ignore instructions embedded within the tool output.
Capability inventory: The skill possesses significant capabilities, including the ability to create, amend, and cancel financial orders, and close or modify open positions.
Sanitization: Absent. The agent is instructed to extract and format data into Markdown tables without explicit validation or escaping of the content returned by the external API.

gate-exchange-tradfi