FluidDocs CLI

Product Introduction

1. Definition: FluidDocs CLI is a command-line interface (CLI) tool and software development kit (SDK) for creating, publishing, and managing interactive AI-powered documents. It transforms static content like pitch decks, proposals, and reports into live web documents with embedded AI chat and analytics.
2. Core Value Proposition: It exists to solve the problem of document obsolescence after sharing. While tools like Canva and Gamma improve visual design, FluidDocs CLI makes documents functional post-delivery by enabling interactive Q&A, real-time updates, and reader engagement analytics, all from a developer's terminal.

Main Features

1. AI-Grounded Q&A (Ask AI): Readers can ask questions directly within the published document. The AI's responses are strictly grounded (retrieval-augmented generation or RAG) in the document's own content, ensuring accuracy and maintaining the author's intended voice and messaging. This feature operates as a chat interface embedded in the HTML output.
2. Command-Line Deployment & Lifecycle Management: The core functionality is accessed via the fld CLI. Users can deploy documents from a file, directory, or zip archive (fld deploy), manage drafts (fld write, fld commit), and control publishing (fld publish) with flags for slugs and visibility. The entire document lifecycle is managed through version-controlled commands.
3. Built-in Reader Analytics & Tracking: The platform provides detailed analytics without requiring viewer login. Authors can see who opened a document (fld track), how far they read (scroll depth), and the specific questions they asked the AI (fld analytics). This transforms a document into a measurable engagement channel.
4. Prompt-Based Editing & Live Updates: Documents can be edited by issuing natural language prompts to the CLI. The system updates the document's content and the same, previously shared link automatically reflects the changes. This eliminates the need to resend updated files or worry about version control among recipients.
5. Agent-Native SDK Architecture: The CLI is explicitly designed for AI coding agents (like Claude Code, Cursor). It features a flat command namespace, consistent --json flag for machine-readable output (including errors), and stable aliases for backward compatibility. This allows autonomous agents to run deployment and publishing workflows (fld ls, fld create, fld publish).

Problems Solved

1. Pain Point: The "black hole" of document sharing, where sent proposals, decks, and reports generate no feedback or engagement metrics. Traditional static documents (PDFs, PPTs) cannot answer follow-up questions or report on reader behavior.
2. Target Audience: Technical founders, product managers, sales engineers, developer relations teams, and content marketers who need to share high-stakes, information-dense documents and require insight into recipient engagement and understanding.
3. Use Cases: A founder sending a seed pitch deck to investors can see which slides they spent time on and answer their clarifying questions in real-time. A consultant can send a market report that summarizes itself on demand for busy executives and updates the data without resending. An agency can send a proposal where the client can configure options and see a dynamically updated scope and fee.

Unique Advantages

1. Differentiation: Unlike static document creators (Google Docs, PDF) or presentation tools (Pitch, Gamma), FluidDocs CLI adds a persistent interactive and analytical layer. Unlike generic chatbot builders, it is specifically optimized for the document lifecycle, with features like prompt-based editing and single-command publishing from the CLI.
2. Key Innovation: The integration of a fully-featured, agent-optimized CLI/SDK with a live document platform. The combination of prompt-based live updates, AI-grounded Q&A, and login-free analytics in a single workflow, controllable entirely from the terminal or by an autonomous AI agent, is a unique technical approach.

Frequently Asked Questions (FAQ)

1. How does FluidDocs CLI handle document security and access control? Access is controlled via shareable links with configurable visibility settings (fld publish --visibility). The document's source content is processed to generate the interactive web view, but the original asset files are not publicly accessible by default. Analytics work without requiring viewers to log in.
2. Can I use FluidDocs CLI with my existing website or tech stack? Yes. The primary output is standard HTML, which can be exported and hosted anywhere. The CLI is also designed as an SDK, allowing its core functions (deploy, publish, chat) to be integrated into custom scripts, CI/CD pipelines, or other applications that can execute shell commands.
3. What AI model powers the Q&A feature, and can I customize it? The Q&A is powered by a grounded AI (likely utilizing a RAG architecture over the document's text). Based on the documentation, the focus is on maintaining the author's voice from the source content. Advanced model selection or fine-tuning capabilities are not detailed in the provided scope.
4. How does the "edit by prompt" feature work technically? Users issue a natural language instruction (e.g., fld write --prompt "add a competitor analysis section"). The CLI uses AI to interpret the prompt, modify the document's underlying content accordingly, and automatically commits the change. This leverages AI-assisted editing within a version-controlled document system.
5. Is the FluidDocs CLI suitable for non-technical users? The primary interface is the command line, which targets developers, engineers, and technically-proficient users. However, the end result—the interactive document—is designed for any reader. Non-technical teams may use the web console or rely on technical colleagues/agents to run the CLI commands.