The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation of a global Node.js package @jackwener/opencli to provide the core functionality of bridging the Discord desktop application to the terminal. This involves downloading and executing code from an external repository to manage local application interactions.
[DYNAMIC_CONTEXT_INJECTION]: The SKILL.md file employs a dynamic command execution block using the !command syntax to check for the presence of opencli and its connection status. This mechanism executes the shell command (command -v opencli && opencli discord-app status 2>&1 | head -5 && echo "READY" || echo "SETUP_NEEDED") automatically when the skill is loaded by the agent to determine the environment state.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from external Discord servers and search results, creating a potential surface for indirect prompt injection.
Ingestion points: Data enters the agent's context through Discord messages fetched via opencli discord-app read and opencli discord-app search as described in SKILL.md and references/commands.md.
Boundary markers: The instructions do not specify any delimiters or safety markers to help the agent distinguish between its instructions and the content of the retrieved messages.
Capability inventory: The skill environment supports shell command execution, including the installation of packages and execution of local CLI tools.
Sanitization: There is no evidence of sanitization or filtering applied to the messages before they are processed by the agent.

discord-reader