The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill instructs the agent to prompt the user for a sync token (API key) and user ID for service authentication.
[DATA_EXPOSURE]: The skill accesses and writes to sensitive configuration files located in the user's home directory (~/.claude-mem/settings.json) to persist authentication secrets.
[COMMAND_EXECUTION]: The skill executes a local Node.js hook script (cmem-hook.mjs) to verify the installation and configuration status.
[INDIRECT_PROMPT_INJECTION]: The skill processes a 'Connect blurb' provided by the user to extract values, which serves as an ingestion point for potentially untrusted data.
Ingestion points: Connect blurb (SKILL.md)
Boundary markers: Absent
Capability inventory: Local file system writes to plugin and home directory config files, subprocess execution of internal Node.js scripts.
Sanitization: Absent

mem-setup