What problem does it solve? Creating a well-structured agent company package requires understanding the Agent Companies specification, designing an org chart with reporting lines, defining workflows, and writing many interlinked files. This Skill guides the AI through interviewing the user, reading the spec, and generating a complete, spec-compliant company package either from scratch or from an existing git repository. ## Core Features & Use Cases - Two Creation Modes: Build a company from scratch via a structured interview, or analyze an existing git repo (URL, local path, or tweet) and wrap its skills and structure into a company. - Spec-Compliant Package Generation: Produces COMPANY.md, AGENTS.md, TEAM.md, PROJECT.md, TASK.md, SKILL.md files plus a .paperclip.yaml vendor extension, all following the agentcompanies/v1 schema with correct slugs, reporting structures, and source references. - Workflow-Aware Agent Design: Infers the right organizational pattern (pipeline, hub-and-spoke, collaborative, or on-demand) and writes agent instructions that define handoffs, triggers, and execution contracts. - Use Case: A user points the AI at a GitHub repo containing several skills and asks to turn it into a company. The Skill analyzes the repo, proposes a lean agent roster, references the existing skills by commit SHA instead of copying them, and outputs an importable package with README and LICENSE. ## Quick Start Ask the AI to create an agent company from this repository or from a description of the team you want, then answer its interview questions to confirm the agents and workflow.
