The Agent Skills Directory

[COMMAND_EXECUTION]: The skill requires the agent to execute shell commands using the opencode CLI tool to verify model availability (opencode models), debug settings (opencode debug config), and invoke the AI model (opencode run).- [EXTERNAL_DOWNLOADS]: The skill references and links to documentation on opencode.ai for configuration and permission guidelines. It also interacts with remote AI models via the Z.AI Coding Plan.- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and review code from a repository, which is a potential surface for indirect prompt injection if the files contain malicious instructions. The skill proactively mitigates this risk by instructing the agent to treat output as 'leads' and verify all material claims against the source before taking action.- [DYNAMIC_EXECUTION]: The instructions provide a pattern for using Python's subprocess.Popen to manage process execution, process groups, and timeouts for the CLI tool.

glm-review