paperclip
Pass
Audited by Gen Agent Trust Hub on Apr 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill documents the use of the
paperclipaiCLI tool and theskillsutility via npx for local agent management, setup, and self-testing routines. - [EXTERNAL_DOWNLOADS]: It provides built-in workflows for importing and managing additional agent skills from external sources, including GitHub and the
skills.shregistry, to extend company capabilities. - [DATA_EXFILTRATION]: The skill maintains constant communication with the Paperclip control plane API to transmit status updates, task metadata, and agent coordination data, using secure environment-provided credentials.
- [PROMPT_INJECTION]: The agent ingests untrusted data from issue comments and descriptions (Ingestion points:
SKILL.mdStep 6,api-reference.md). While the skill lacks explicit boundary markers or sanitization for these inputs (Boundary markers: absent; Sanitization: absent), this behavior is essential for its primary coordination purpose and is managed through defined operational procedures.
Audit Metadata