The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute npx claude-mem commands and a Node.js script to manage service operations and configuration.
Evidence: npx claude-mem backup status, npx claude-mem restart, and `node
<<'EOF' ... EOFblocks inSKILL.md`.
[CREDENTIALS_UNSAFE]: The skill reads and modifies ~/.claude-mem/settings.json, which is explicitly stated to contain credentials.
Evidence: The skill notes: "Never print the file's contents: it holds credentials." and implements fs.chmodSync(file, 0o600) in the configuration script to restrict access.
[DYNAMIC_EXECUTION]: The skill generates and executes a Node.js script at runtime to modify the local filesystem and configuration settings.
Evidence: The HEREDOC `node
<<'EOF'block in Step 2 ofSKILL.md` which parses and writes JSON configuration.
[INDIRECT_PROMPT_INJECTION]: The skill processes output from external CLI commands which could potentially contain untrusted data if the underlying database or filenames were manipulated.
Ingestion points: Output from npx claude-mem backup status and npx claude-mem backup list is read by the agent.
Boundary markers: Absent; no specific delimiters are defined for the tool output.
Capability inventory: The skill has access to Bash (command execution) and Read (file system access).
Sanitization: Absent; the skill instructions do not specify validation or sanitization of the CLI output before the agent processes it.

backup-setup