The Agent Skills Directory

[COMMAND_EXECUTION]: The skill operates primarily by executing shell commands through the paperclipai CLI to inspect company dashboards, agent health, and issue status.
[EXTERNAL_DOWNLOADS]: The documentation provides instructions to install the paperclipai package via npm and to run the @bbengamin/paperclip-mcp-server using npx. These are vendor-owned tools necessary for the skill's functionality.
[CREDENTIALS_UNSAFE]: The skill references local configuration and authentication files located at ~/.paperclip/context.json and ~/.paperclip/auth.json. The instructions include explicit security warnings to never print bearer tokens or API keys in logs or chat sessions.
[PROMPT_INJECTION]: The skill processes external data such as issue descriptions and activity logs from the Paperclip platform, which presents a surface for indirect prompt injection.
Ingestion points: Output from paperclipai issue list, agent list, and activity list commands.
Boundary markers: The workflow requires structured reporting and distinguishing between different run states (e.g., queued vs. failed).
Capability inventory: The skill has the ability to execute shell commands and perform network requests via the Paperclip REST API.
Sanitization: The skill enforces a strict read-only policy by default and requires explicit operator approval before recommending or performing any state-changing mutations.

paperclip-monitor