The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses dynamic context injection (the ! command syntax) in SKILL.md to automatically execute a shell command that reads the preferences.json file upon loading. This injects raw secrets and API keys directly into the agent's context.
[CREDENTIALS_UNSAFE]: The skill is designed to target and read a sensitive configuration file located at ~/.claude/plugins/data/ops-ops-marketplace/preferences.json to access stored integration tokens.
[DATA_EXFILTRATION]: The skill reads local credentials and uses curl to transmit them to external API endpoints belonging to services such as Stripe, Slack, and Klaviyo for verification.
[EXTERNAL_DOWNLOADS]: The skill utilizes the npx command to fetch and execute the @dopplerhq/mcp-server package from the npm registry during a smoke test.
[REMOTE_CODE_EXECUTION]: The integration tests involve downloading and running external code via npx, which executes remote logic within the local environment.

ops-settings