The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from a local configuration file that could potentially be modified by other processes to influence agent behavior.
Ingestion points: ~/.mem0/settings.json is read using shell commands and the Read tool.
Boundary markers: The instructions do not define explicit delimiters or 'ignore' instructions for the content read from the settings file.
Capability inventory: The skill has the ability to execute shell commands (grep, tr) and write to the filesystem via the Write tool.
Sanitization: The skill employs specific regular expression patterns to extract only the default_scope key, which reduces the risk of arbitrary instruction injection from other fields in the JSON file.
[COMMAND_EXECUTION]: The skill utilizes local shell commands to parse configuration data.
Evidence: In SKILL.md, a shell pipeline using grep and tr is used to extract the current scope value from ~/.mem0/settings.json.

mem0-scope