The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to create and run temporary verification scripts in tempfile.gettempdir() using a hermes-verify- prefix. This pattern of generating and executing code based on runtime state can be exploited if the inputs to those scripts are not strictly sanitized.
[INDIRECT_PROMPT_INJECTION]: The workflow involves reading today's daily note (specifically ## Log and ## Wins sections) and Todoist items to generate tomorrow's note. This represents an ingestion point for untrusted data that could contain malicious instructions designed to influence the agent's behavior during the wrap-up process.
[PRIVILEGE_ESCALATION]: The provided script scripts/atomic-note-replace-verify.sh utilizes chflags uchg and chflags nouchg on macOS to set and unset the system immutable flag. While intended for file integrity, this demonstrates use of system-level file attributes to lock resources.
[REMOTE_CODE_EXECUTION]: The instructions mandate using a specific shell script (scripts/atomic-note-replace-verify.sh) for atomic file operations. While the script is provided locally within the skill, the pattern of directing the agent to execute specific shell scripts for file system recovery increases the risk if the script paths or arguments are manipulated.

daily-note-wrapup