The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use shell commands including pwd, tr, and grep to dynamically resolve the path to the agent's memory directory and search for duplicate entries within MEMORY.md.
[PROMPT_INJECTION]: The skill facilitates an Indirect Prompt Injection surface by allowing arbitrary, unsanitized user input to be written to a persistent memory file (MEMORY.md) that the agent reads at the start of every session.
- Ingestion points: User-supplied text provided to the /remember command is the primary entry point for untrusted data into the persistent context.
- Boundary markers: Missing. Content is appended to the memory file as a simple Markdown list item (- {{fact}}) without explicit delimiters or instructions to the agent to treat the content as data rather than instructions.
- Capability inventory: The skill utilizes file reading (grep) and file writing capabilities to manage the memory storage.
- Sanitization: None. The skill does not perform any validation, filtering, or escaping of the user-provided text before persisting it to the filesystem.

remember