The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local Node.js script located at ~/.codex/supermemory/status.js to retrieve connection information.
[COMMAND_EXECUTION]: The skill explicitly instructs the agent to request escalated shell execution on Windows if sandbox execution is likely to fail. This pattern is designed to circumvent security boundaries established by the sandbox environment to facilitate access to files in the user's home directory.
[DATA_EXFILTRATION]: The skill accesses the ~/.codex/supermemory/ directory, which is a sensitive location used for storing application credentials and configurations. While this is a vendor-owned resource for the 'supermemoryai' author, it constitutes access to sensitive configuration data.
[PROMPT_INJECTION]: The skill processes output from a local script and incorporates it into the response, creating an indirect prompt injection surface.
Ingestion points: Data is ingested from the output of the node ~/.codex/supermemory/status.js command.
Boundary markers: No delimiters or protective instructions are used to separate the ingested script output from the agent's instructions.
Capability inventory: The skill is permitted to execute commands via the Bash(node:*) tool.
Sanitization: No explicit validation or sanitization is performed on the script output before it is processed by the agent, although the skill instructions recommend preserving the script's own credential masking.

supermemory-status