The Agent Skills Directory

[COMMAND_EXECUTION]: The skill requires the execution of shell commands to search and list files within the local directory structure.
Evidence: Instructions include using grep, ls, findstr, and Select-String depending on the host operating system to locate information in the memory/ directory and MEMORY.md file.
Context: These commands are limited to local file searches and do not involve network operations or administrative privileges.
[INDIRECT_PROMPT_INJECTION]: The skill establishes an attack surface for indirect prompt injection by reading files that might store untrusted external data.
Ingestion points: Data enters the agent's context through the native read_file tool when accessing MEMORY.md and daily logs in the memory/ directory.
Boundary markers: The skill does not specify the use of delimiters or 'ignore' instructions when reading retrieved memory content.
Capability inventory: The agent has the ability to write to the filesystem (write_file, edit_file) and execute shell search commands.
Sanitization: No sanitization or validation logic is defined for data before it is written to or read from the memory files, though there is a manual guideline for the agent to avoid recording sensitive credentials.

nano-memory