The Agent Skills Directory

[SAFE]: The skill is designed to interact with the author's own infrastructure (Paperclip) using dedicated tools like paperclip_connection, paperclip_create_task, and paperclip_read_task. No unauthorized network communication or suspicious third-party dependencies were found.
[SAFE]: The instructions do not contain any prompt injection attempts, obfuscated content, or persistence mechanisms. The delegation logic follows standard operational patterns, such as verifying connections and handling disconnected calls through task inspection rather than repeating actions.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits a standard ingestion surface by incorporating user-provided context into task descriptions for downstream agents. This is an intended functional requirement for task delegation and does not contain evidence of malicious intent.
Ingestion points: User-provided context, objectives, and deliverables described in SKILL.md.
Boundary markers: Absent; the skill relies on clear scope resolution with the user.
Capability inventory: Limited to platform-specific task management tools (paperclip_create_task, paperclip_read_task) as seen in SKILL.md.
Sanitization: Not explicitly defined within the markdown instructions.

delegate-work