The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs users to install an external package @jackwener/opencli from NPM and a supplementary plugin from a GitHub repository (github:himself65/finance-skills/hyperliquid).
[COMMAND_EXECUTION]: The skill relies on shell commands to interact with the Hyperliquid API via the opencli tool. It also implements a dynamic context injection pattern in SKILL.md to run an environment status check (using command -v and opencli markets) when the skill is initialized.
[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests and processes market data from an external API (api.hyperliquid.xyz).
Ingestion points: Market data is retrieved from https://api.hyperliquid.xyz/info as described in SKILL.md and references/commands.md.
Boundary markers: The skill does not specify the use of delimiters or clear indicators to separate external API content from the agent's instructions.
Capability inventory: All scripts and instructions rely on shell command execution via the opencli utility as documented in SKILL.md and references/commands.md.
Sanitization: There is no evidence of data sanitization or validation performed on the API responses before they are presented to the AI agent.

hyperliquid-reader