doc-coauthoring
A structured workflow for co-authoring documentation, technical specs, and proposals, guiding users through context gathering, collaborative refinement, and reader verification.
Introduction
The doc-coauthoring skill is designed to transform unstructured ideas into professional-grade documentation through a systematic, three-stage collaborative process. It acts as an intelligent partner for technical writers, engineers, and product managers tasked with creating complex documents such as PRDs, RFCs, design documents, or project proposals. By managing the flow of information, the agent ensures that the final output is context-rich, structurally sound, and tailored for its intended audience.
-
Facilitates a three-stage workflow: Context Gathering, Refinement & Structure, and Reader Testing to maximize document clarity.
-
Automatically identifies and prompts for essential metadata, including document type, target audience, desired impact, and template requirements.
-
Employs an iterative brainstorming approach, generating multiple options per section to allow users to select, combine, or refine content surgically.
-
Includes a dedicated Reader Testing phase, utilizing a fresh, context-free agent session to identify logical gaps or blind spots that might confuse future human readers.
-
Offers integration support to fetch context directly from tools like Google Drive, Slack, SharePoint, or other MCP servers, and provides guidance on enabling connectors if external tools are currently unavailable.
-
Trigger this skill by mentioning specific tasks like writing a spec, drafting a proposal, or creating a decision document.
-
Provide context freely as a stream-of-consciousness dump; the agent will parse this and ask targeted, numbered follow-up questions to fill critical gaps.
-
Use the provided structure to iterate through individual sections rather than attempting to draft the entire document at once, focusing on high-uncertainty areas first.
-
If your document contains images, be prepared to upload them during the workflow, as the agent can generate descriptive alt-text to ensure accessibility for future readers using AI.
-
The workflow is highly adaptable; while it enforces a structured path, users may choose to work freeform if the situation requires flexibility or if the document scope is limited.
Repository Stats
- Stars
- 125,607
- Forks
- 14,716
- Open Issues
- 785
- Language
- Python
- Default Branch
- main
- Sync Status
- Idle
- Last Synced
- Apr 29, 2026, 05:26 AM