The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill utilizes the paperclipai CLI and the @bbengamin/paperclip-mcp-server package, both of which are vendor-provided resources associated with the skill author. These are installed or executed via standard package managers (npm/npx).
[COMMAND_EXECUTION]: The skill performs shell commands using the paperclipai CLI and curl for API interactions. All operations that modify state (mutations) are explicitly gated by an 'Operator Approval Boundary,' ensuring the human operator reviews and approves changes before execution.
[DATA_EXFILTRATION]: While the skill manages API authentication, it includes strict directives to never print or log bearer tokens. It encourages the use of environment variables and established local auth stores (~/.paperclip/auth.json) for credential management.
[PROMPT_INJECTION]: The skill ingests data from external wiki pages and previously generated summaries. The potential for indirect prompt injection is mitigated by the requirement for the agent to present proposed mutations to the operator for manual approval before updating the Paperclip control plane.
[SAFE]: The skill implements robust security practices for an administrative tool, focusing on transparency, human oversight, and the protection of sensitive authentication data.

growth-record-strategy