docs: Marketing one-pager

Author: zah

docs/marketing-one-pager.md

@@ -0,0 +1,61 @@
+### **Unlocking Developer Productivity: The Agents-Workflow Advantage**
+
+**For Marketing & Business Development Teams**
+
+#### **Executive Summary**
+
+Agents-Workflow is a comprehensive, highly-opinionated platform designed to maximize the efficiency, security, and effectiveness of AI coding agents. It provides a powerful command-line interface and a robust backend that seamlessly integrates AI into the software development lifecycle. By addressing the critical pain points of existing cloud-based systems—such as slow start-up times, difficult environment replication, and the inability to precisely guide "runaway" agents—Agents-Workflow empowers developers to delegate complex tasks to AI with confidence, freeing them to focus on high-level architecture and innovation.
+
+---
+
+#### **The Agents-Workflow Difference: Key Benefits**
+
+While cloud-based AI coding assistants offer a glimpse into the future, Agents-Workflow delivers a practical, powerful, and polished solution for today.
+
+* **Parallel, Autonomous Workstreams:** Launch multiple, fully sandboxed AI agents that can work on different tasks concurrently. Our system doesn't require constant developer supervision or confirmation for routine commands, allowing developers to remain productive while the agents work towards delivering a high-quality, test-passing Pull Request.
+* **Precision Intervention with Agent Time-Travel:** Go beyond live monitoring. Our unique session recording and branching functionality allows developers to rewind an agent's entire work session, inspect the filesystem at any precise moment, and branch off with new instructions to correct the agent's course.
+* **Instantaneous, Isolated Workspaces:** Leveraging Copy-on-Write (CoW) filesystem snapshots like ZFS and Btrfs, we can create fully isolated, writable snapshots of a project's entire working tree in milliseconds. This means agent start-up is nearly instantaneous, and incremental builds work perfectly, dramatically reducing agent waiting time.
+* **Seamless Environment Replication:** With first-class support for `devcontainers`, Nix, and Spack, we eliminate environment configuration drift. Developers can define their exact development environment once, ensuring local agents, on-premise clusters, and cloud agents all operate with identical tooling, dependencies, and configurations.
+* **Auditable & Collaborative History:** Every task is recorded, creating a fully auditable trail of all work performed. This history can be stored either directly in the project's version control system or within the on-premise company backend, accessible through a dedicated web portal.
+
+---
+
+#### **Addressing the Pains of Modern AI-Assisted Development**
+
+We directly solve the most common frustrations developers and organizations face with existing AI coding platforms.
+
+##### **Problem 1: The "Setup Tax" and Prohibitively Slow Start-Up Times**
+Cloud agents like Codex and Jules are ineffective without a perfectly replicated development environment. Developers spend significant time writing and debugging complex setup scripts. Worse, these scripts must be executed from scratch for every new task, making agent start-up painfully slow and expensive.
+
+**Our Solution: A Spectrum of Instant-On Environments**
+Agents-Workflow offers a range of start-up options to fit any scenario, all managed through a unified, seamless user experience. We support and augment existing cloud agent environments to ensure a consistent workflow, regardless of the underlying runner.
+* **Local & On-Premise:** Launch agents locally or on an on-premise cluster, leveraging a full development environment. With our snapshot technology, a new, fully-provisioned agent workspace can be started in milliseconds.
+* **Cloud:** Utilize `devcontainer` configurations to enable efficient launch of cloud agents, benefiting from existing infrastructure like GitHub Codespaces while ensuring perfect environmental consistency.
+
+##### **Problem 2: The "Runaway Agent" and Lack of Precise Control**
+AI agents are fast, but they can quickly go down the wrong implementation path or give up on meeting all stated requirements when they face difficulties. Trying to interject and correct them during a live session is often difficult and imprecise, leading to wasted time and suboptimal results.
+
+**Our Solution: Precise Guidance with Agent Time-Travel**
+Our platform's most powerful feature provides a complete solution.
+* **Session Recording:** Every agent session, including all terminal I/O, is recorded.
+* **Filesystem Snapshots (`FsSnapshot`):** At key moments (e.g., before and after a command runs), we take a near-instantaneous, low-overhead snapshot of the entire workspace.
+* **Time-Travel & Branching (`SessionBranch`):** A developer can pause the session playback at any point, explore the exact state of the code from that moment, and if a correction is needed, create a "session branch." This launches a *new* agent session from that precise point in time with a new, corrective instruction, creating a parallel timeline without disrupting the original. This is the ultimate tool for guiding agents effectively.
+
+---
+
+#### **Deep Dive: Key Features**
+
+* **High-Performance Workspaces via Filesystem Snapshots:**
+    The core of our speed and isolation advantage comes from our intelligent use of modern filesystems. The system automatically detects and uses the best available technology:
+    * **Linux:** ZFS and Btrfs for instant, zero-cost snapshots and clones.
+    * **Cross-Platform Fallback:** For other environments (including macOS and Windows), we can fall back to a high-performance, in-memory Copy-on-Write (CoW) filesystem that runs in user space, avoiding the need for a Linux VM while still providing significant speed benefits. The use of a Linux VM remains a supported option for maximum compatibility.
+
+* **Unified Development Environments (Devcontainers & Nix):**
+    By embracing industry standards like `devcontainers`, we solve the "it works on my machine" problem for AI agents. This ensures that every agent, whether running locally or in a private cloud, has access to the exact same dependencies, linters, and testing frameworks as the human developers, leading to higher-quality, more reliable results.
+
+#### **Business & Organizational Impact**
+
+* **Amplify Developer Productivity:** Drastically reduce time spent on environment setup and agent supervision. Enable developers to run more AI-driven experiments in parallel, accelerating innovation.
+* **Improve Code Quality and Consistency:** Agents that can run the full suite of project tests and linters in a correct environment produce vastly superior code. The Time-Travel feature allows for meticulous review and course correction, ensuring high standards are met.
+* **Enhance Knowledge Sharing and Onboarding:** The recorded history of tasks and coding sessions creates a valuable and durable knowledge base. New hires and team members can learn by reviewing how specific problems were prompted, approached, and solved by both senior engineers and AI agents.
+* **Develop Actionable KPIs:** The transparent, auditable trace of all agent activity allows organizations to develop meaningful KPIs for AI-assisted development, track productivity gains, and identify best practices to scale across the entire team.