The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructions do not contain any attempts to bypass safety filters, override system constraints, or extract system prompts. It uses standard instructional language to define its operational logic.
[DATA_EXFILTRATION]: No network operations (such as curl, wget, or fetch) or access to sensitive local files (such as credentials or SSH keys) are present. The skill only interacts with the user within the chat context.
[REMOTE_CODE_EXECUTION]: The skill does not perform any remote code downloads, package installations, or dynamic script execution. It relies entirely on natural language processing.
[COMMAND_EXECUTION]: No shell commands or subprocess calls are present. The only 'commands' mentioned are internal skill triggers like /sort and /compose, which are standard navigational references within the agent's environment.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-provided notes and ideas. While it lacks explicit boundary markers to delimit untrusted user content from instructions, the skill does not possess any dangerous tools (e.g., network access or file writes) that could be exploited via malicious input. The risk is limited to the output text itself.

sequence