Composable workflows for the agent stack.
Pre-built and custom playbooks turn signals from Lens and data from Prism into briefs, audits, FAQs, and content, on a schedule or on demand.
Templates that work day one
Competitor audits, AI FAQ generators, content briefs, video repurposers, 24 templates tuned for AI visibility.
Triggered or scheduled
Run on a cron, fire on a Lens alert, or kick off manually. Outputs land in Slack, email, or your CMS.
Auditable by default
Every run logs sources, prompts, and outputs. Show your CMO why the agent recommended what it did.
Trigger → source → transform → output.
Every playbook is a graph of four node types. Compose them in the UI, or drop to YAML when you need the full schema.
Start a run
Cron schedule, Lens alert, webhook, manual kick-off, or chained from another playbook.
Pull the data
Lens queries, Prism SKUs, your CMS, competitor sites, Reddit, Trustpilot, whatever the playbook needs.
Run the model
Claude, GPT-4.1, or Gemini, with structured output, tool calls, and deterministic caching.
Ship the work
Slack, Notion, Linear, email, webhook, Contentful, or straight back into Prism as enriched attributes.
24 playbooks, working on day one.
Every template is versioned, auditable, and forkable. Start from one of these, or build from an empty graph. Here are nine of the most-run:
Competitor citation audit
Scrape agent responses for your category, rank competitor mentions, flag new entrants, and deliver a CMO-ready diff.
AI FAQ generator
Monitor long-tail prompt clusters, detect recurring questions, generate structured FAQ blocks, push to CMS.
Content brief from citations
Pick the top 10 pages agents cite. Generate a competitive brief for each, assigned to a writer in Linear.
Video to AI-ready transcript
Pull YouTube and TikTok content, transcribe, extract claims, push structured Q&A back into Prism for agent retrieval.
Sentiment defense brief
Fires when Lens detects a sentiment drop. Pulls the citation sources, drafts response content, assigns to PR and support.
Attribute enrichment sweep
Scans Prism for SKUs missing material, fit, or certification fields. Generates and validates copy, pushes back to catalog.
Morning visibility digest
One email. Three insights ranked by business impact. Every morning at 08:00 local, in your head of digital's inbox.
Board-ready AI share report
Quarterly deck: visibility trend, competitive position, attributed revenue uplift, and next-quarter priorities.
New market readiness audit
Launching in a new region? Runs 400 localised prompts, maps competitive landscape, scores feed readiness.
Ships to where your team already works.
Every playbook can fan out to any combination of destinations. Add webhooks to hit anything else.
Every run, every prompt, every source.
Playbooks keep the full execution trace for 18 months, models used, prompts sent, sources retrieved, outputs produced. Enterprise tier adds SOC-2 export and signed audit bundles.
Which model ran, with what system prompt, at what temperature, and which version of the playbook produced the output.
Every URL, SKU, and Lens citation that informed a run, with timestamps and the raw content that was retrieved.
Optional human-in-the-loop step for CMS pushes. Diff the current and proposed content, approve in Slack, ship on approval.
We replaced a sixteen-step manual Monday reporting ritual with one Playbook. Our CMO gets better insight, earlier, without touching a spreadsheet.
Pairs with Lens and Prism.
Playbooks read from Lens signals and Prism catalogs, and write back to both. The three products are designed to loop.
The source of most triggers.
Every Lens alert can fire a playbook. Rank drops, sentiment shifts, new prompt clusters, all become work automatically.
Explore LensThe most common destination.
Generated FAQs, enriched attributes, and canonical copy flow back into Prism, instantly improving what agents retrieve.
Explore Prism