What problem does it solve? AI agents often hand over work with no record of what they based it on, making outputs impossible to verify or audit. This Skill enforces a standard agent folder structure and requires every delivered piece of work to cite the files and conversation lines it drew on. ## Core Features & Use Cases - Standardized Agent Folder Structure: Defines a three-part layout under the agent directory: a brief.md file, a skills/ folder, and a documents folder for specs, data, and past decisions. - Structured Brief Template: The brief.md file contains four fixed sections — Identity, Constraints, Read first, and Inputs/Outputs — so each agent's role, limits, and expectations are explicit. - Mandatory Citation of Sources: Every output ends with a short "Used:" list naming the files and conversation lines it relied on, with anything uncheckable marked as unverified. - Use Case: A drafting agent produces a report from documents/spec.md and prior conversation; the delivered report ends with a three-line citation block so the reviewer can verify each claim against its source. ## Quick Start Set up my agent's folder with a brief.md following the four-section structure and make sure every output it hands in ends with a Used citation list.
