The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation of the @jackwener/opencli package via NPM and the manual download of a browser extension from a GitHub repository (https://github.com/jackwener/opencli/releases). These sources are not recognized as trusted organizations or well-known services.
[COMMAND_EXECUTION]: The skill instructs the agent to execute the opencli command with arguments (site names, subcommands, and flags) derived from user input. This allows the agent to interact with various external sites through a CLI interface. The skill includes warnings against using mutation-style commands (e.g., post, delete) to mitigate risk.
[DYNAMIC_CONTEXT_INJECTION]: In SKILL.md, the skill uses the !command`` syntax to execute shell commands (command -v opencli and opencli doctor) at load time. This is used to verify the tool's installation and configuration status before the agent processes the skill instructions.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to fetch and process content from a wide array of external platforms (Reddit, Bloomberg, arXiv, etc.). This creates a vulnerability where malicious instructions embedded in the retrieved web data could attempt to manipulate the agent's behavior.
Ingestion points: Untrusted data enters the agent context through opencli command outputs in SKILL.md and the reference files.
Boundary markers: The skill does not define specific delimiters or provide instructions to the agent to ignore potentially malicious embedded content in the fetched data.
Capability inventory: The skill utilizes shell command execution and network access (via the opencli tool).
Sanitization: There is no evidence of automated sanitization or filtering of the data returned by the external tool before it is processed by the agent.

opencli-reader