# Playbooks Playbooks let you describe work as natural-language instructions, attach context like a Brand Kit, and run the process end to end while keeping reviewers in the loop. A Playbook is made of: * **A doc-style Playbook** that describes what the agent should do. * **Inputs** that the Playbook accepts at run time, such as text, numbers, files, or a Brand Kit. * **Sections** that contain instructions, tool calls, and Human Review checkpoints. * **Tools** that the agent can use, including SEO Research, Web Research, Image and Video, MCP Connectors, and more. * **Artifacts** that the Playbook produces, such as Markdown, HTML, JSON, CSV, PNG, JPG, or GIF files. * **Triggers** that determine how a Playbook starts, including Schedule, Webhook, Monitor, and AEO Insight when enabled. {% hint style="info" %} Playbooks use the same underlying engine as Workflows, including Brand Kits, Knowledge Bases, SERP analysis, LLM models, and the AirOps tool catalog. The authoring and review experience is different. {% endhint %} ## Key concepts | Term | Definition | | ------------ | --------------------------------------------------------------------------------------------------------------------------------- | | Playbook | The agent definition with Inputs, Sections, Tools, Artifacts, and a Trigger. Running a Playbook produces a Session. | | Section | A unit of work in a Playbook. Each Section can include instructions, tool calls, and a Human Review block. | | Input | Information you provide when a Playbook runs, such as a keyword, target URL, file upload, or Brand Kit. | | Tool | An action the agent can take, from web search and SEO research to MCP Connectors like Slack, Notion, and GitHub. | | Artifact | A persistent output the Playbook produces, such as a blog post, image, analysis, or structured data. | | Trigger | A condition that starts a Playbook automatically. Trigger types include Schedule, Webhook, Monitor, and AEO Insight when enabled. | | Session | A single Playbook run. Each Session produces Artifacts and logs the work the agent performed. | | Inbox | The workspace-level queue where reviewers see pending Human Review steps and Session outputs that need attention. | | Run History | The record of every Session a Playbook has produced. | | Human Review | A pause point where a person must approve or edit the agent's work before it continues. | | Quill | The AirOps agent captain that helps you build and edit Playbooks from natural-language prompts. | | Subagent | A specialized agent a Playbook can call, such as Writing, Brand Compliance, or Brand Kit. | | Memory | Persistence across Sessions, used when a Playbook needs to remember what it has already surfaced or completed. | ## Playbooks vs. Workflows Playbooks are the default place to start for new action-oriented content work in AirOps. Workflows remain available for lower-level visual automation, existing workflow use cases, and cases where you need explicit step-by-step control. | Dimension | Workflows | Playbooks | | --------------- | ----------------------------------------------------------------- | --------------------------------------------------------------------------------------------- | | Authoring | Step-by-step visual builder. You wire each step explicitly. | Natural-language Playbook. You describe the goal and the agent figures out execution. | | Inputs | Input fields that you define and map. | Typed Inputs, including Text, Number, File, and Brand Kit. | | Review | Inspect output manually, then trigger the next step. | Human Review blocks with reviewer assignment and an Inbox queue. | | Iteration | Edit individual steps in the visual builder. | Use Quill, the AirOps agent captain, to edit the Playbook with natural-language instructions. | | Output | Grid cells, webhook payloads, CMS pushes. | Rich Artifacts with deep links, plus the same integration options. | | Triggers | Manual, API, and schedule. | Schedule, Webhook, Monitor, and AEO Insight when enabled. | | Brand grounding | Brand Kit is available but optional. | Brand Kit is a first-class Input. | | Persistence | No memory across runs. | Memory across Sessions supports patterns like avoiding repeated recommendations. | | Engine | Brand Kit, Knowledge Bases, SERP analysis, LLM models, and tools. | Same engine. | ## The two context sources that drive quality Before you build a Playbook, make sure AirOps Insights and your Brand Kit are current. These two context sources shape the quality of every Session. ### Configure AirOps Insights AirOps Insights is the source for AI search visibility data, including tracked prompts, LLM search responses, citations, competitor presence, and content opportunities. When a Playbook researches AEO opportunities or LLM search responses, it should use the data tracked in AirOps Insights rather than treating AI search as generic web research. Cover the topics, personas, regions, and funnel stages you care about. Expand coverage from sources like Google Search Console queries, sales-call language, support tickets, and competitor citation gaps. The common pattern is: AirOps Insights tracks prompts and LLM search responses, a Playbook monitors that data and finds opportunities, then a second Playbook takes action and creates or refreshes the output. ### Keep your Brand Kit fresh Every Playbook can inherit your Brand Kit, so changes to voice, writing rules, audiences, regions, product lines, visual guidelines, and competitor context flow into Playbook output. A stale Brand Kit degrades downstream output. Audit the Brand Kit on a recurring cadence. Refresh it when you launch a product, change positioning, add a new audience, or expand into a new region. Use [Brand Kit Refresh](/context/brand-kit/refresh.md) to monitor live sources and propose Brand Kit edits for human approval. {% hint style="warning" %} Use the new Brand Kit format with Playbooks. Legacy Brand Kits do not work as well with Playbooks and can reduce output quality. {% endhint %} ## Anatomy of a Playbook When you open the Playbook editor, you will see: * A header chip row above the first Section showing Tools, Inputs, Artifacts, and **Add Trigger**. * Top-right actions for **Run Playbook**, **Publish**, and **Run History**. * **Quill, the AirOps agent captain,** as a green floating button in the bottom-right of the editor. * Reference pills inline as you write. Inputs and Brand Kits appear as green pills. Tools appear as gray pills with icons. Type `/` in any Section to open the slash menu. The slash menu surfaces Tools, Inputs, Artifacts, and other references you can insert into your instructions. ## Next steps * [Create a Playbook](/actions/playbooks/create-a-playbook.md) * [Build a Playbook](/actions/playbooks/build-a-playbook.md) * [Run a Playbook](/actions/playbooks/run-a-playbook.md) * [Connect Playbooks to other tools](/actions/playbooks/integrations.md) * [Apply Playbook best practices](/actions/playbooks/best-practices.md) * [Troubleshoot Playbooks](/actions/playbooks/troubleshooting.md) --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://docs.airops.com/actions/playbooks.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.