Docs for AI-assisted reading

OOMOL docs are designed to be read, searched, and summarized by AI. Give the complete docs site to your AI assistant first, let it understand the product structure, commands, tool access paths, and reusable Skills, then ask questions directly.

The Main Path

1. oo-cli

Use oo-cli as the main entry point when you want agents to call real APIs and reusable capabilities.

  • Best starting point for Codex, Claude Code, terminal agents, and other local agent hosts
  • Lets agents call Tools from connected Apps, packages, Skills, and managed custom tools
  • Keeps the shortest path when an existing App, Tool, package, or Skill already solves the job

2. Apps, Tools, and approvals

Use Apps and Tools to describe what people connect and what agents call.

  • Apps are products or services connected to OOMOL, such as Gmail, Slack, GitHub, Notion, Linear, files, databases, or model providers
  • Tools are the callable actions agents use through OOMOL
  • OOMOL keeps account access, permissions, usage, and execution records out of the agent’s raw prompt context

3. Skills

Use Skills when a task is no longer a one-off API call, but a repeatable way to use Tools.

  • Preserve which Tools to call, what context to include, what constraints to follow, and how to check the result
  • Turn useful ways of using Gmail, Notion, GitHub, Slack, Linear, and other Apps into reusable agent guidance
  • Keep agents from rediscovering the same tool use pattern every time a similar task appears

How The Docs Are Organized

  • Apps and Tools: read this first if you need the naming model for connected products and callable agent capabilities
  • oo-cli: start here if you want agents to call Tools from Apps, packages, Skills, and managed custom tools
  • oo-cli Command Reference: use this when you need exact command names, options, and behaviors
  • Support: use this for community, publishing, and related operational topics

User Path And Product Layers

graph TB
    User["User / Agent<br/>wants to use a tool"]

    subgraph UserPath["Direct user-facing entry points"]
        Community["Community / Catalog<br/>discover packages and examples"]
        CLI["oo-cli<br/>search, inspect, and call Tools"]
        Apps["Connected Apps<br/>approved accounts and actions"]
        Skills["Skills<br/>reusable ways to use tools"]
        Records["Records<br/>permissions, usage, files, and logs"]
    end

    User -->|"find an existing capability"| Community
    User -->|"call from a terminal or agent"| CLI
    Community -->|"choose a package"| CLI
    CLI -->|"use approved access"| Apps
    Apps -->|"run a Tool"| Records
    Records -->|"stable method"| Skills
    Skills -->|"reuse next time"| CLI
    Records -->|"task complete"| Done["Task complete"]
  • To find existing capabilities, start from the community or catalog, then use oo-cli to search, inspect, and call Tools from connected Apps, packages, and Skills.
  • When a tool use pattern becomes stable, preserve it as a Skill so agents can reuse the same instructions, constraints, checks, and result shape.
  • Use the command reference when your assistant needs exact CLI syntax instead of a conceptual explanation.