The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes session history to generate a copyable shell command, creating a surface for potential indirect prompt injection if the session history is manipulated.
Ingestion points: The skill reads the active conversation history in SKILL.md.
Boundary markers: No explicit delimiters or ignore instructions are provided for the conversation input.
Capability inventory: The skill executes local commands pwd, scutil, and hostname to gather session metadata.
Sanitization: The instructions include a requirement to shell-quote variables to prevent command injection in the output generated for the user.
[COMMAND_EXECUTION]: The skill utilizes local shell commands to gather system metadata for the resume note.
Evidence: Instructions in SKILL.md specify the use of pwd, scutil --get ComputerName, and hostname.

resume-note