Markup AI

@markup list all available content analysis agents in the 'quality' category so I can pick one for our blog review workflow

Found 4 agents in the 'quality' category: Content Quality Analyzer (checks readability, tone consistency, structural issues), Brand Voice Compliance Checker (validates adherence to style guides), Terminology Validator (flags unapproved terms against your glossary), and Accessibility Auditor (evaluates WCAG compliance for web content). Each agent returns a compact summary here; you'll need to fetch the full agent to see its complete input/output schemas before execution.

This example shows agent discovery — the first step before running any Markup AI analysis. The list operation returns summaries only; you must call GET_AGENT separately to understand what input each agent expects and what structured output it returns. Useful when building content pipelines that need to select the right analyzer dynamically.

@markup run the Brand Voice Compliance Checker agent on this draft using our 'Enterprise Tech' voice profile, then show me which sections failed the check

Workflow initiated (ID: wf_8x3k9m). The agent is analyzing your draft against the 'Enterprise Tech' voice profile, which enforces formal tone, avoids contractions, and requires third-person perspective. Analysis complete: 3 sections flagged. Section 2 uses conversational phrasing ('we've seen'), Section 5 contains a contraction ('don't'), and Section 7 shifts to first-person ('I recommend'). Each violation includes the exact text span and a suggested rewrite aligned with your profile's rules.

This demonstrates asynchronous workflow execution — Markup AI agents run as background jobs, not instant responses. You reference a stored analysis profile (brand voice, style guide, persona) by name to configure the agent's behavior. The workflow ID lets you poll for completion or retrieve results later. Side-effect: the workflow persists in Markup's system and counts toward your plan's execution quota.

@markup fetch the latest 5 completed workflows for the Terminology Validator agent and summarize which unapproved terms appeared most frequently across all runs

Retrieved 5 completed Terminology Validator workflows from the past week. Cross-run analysis shows 'utilize' appeared 12 times (your glossary prefers 'use'), 'leverage' appeared 8 times (flagged as jargon), and 'synergy' appeared 3 times (banned term). These workflows analyzed 23 documents total. The agent's output schema includes term positions and approved alternatives, so you can automate find-and-replace suggestions. Note: workflow results remain accessible for 30 days under your current plan.

This example pairs Markup's workflow history with AI reasoning to surface patterns the MCP alone doesn't compute. Requires LIST_WORKFLOWS filtered by agent name, then GET_WORKFLOW for each result to extract term violations. Useful for editorial teams tracking recurring style issues. Limitation: workflow retention depends on your Markup plan tier, and pagination is cursor-based — you can't jump to arbitrary pages.

When Markup AI makes sense for multi-client style enforcement

A 6-person content agency managing 12 client brands needs to audit blog drafts before publication. Each client has a documented voice guide—one wants Oxford commas, another bans passive voice, a third requires industry jargon. Markup AI's analysis profiles let you encode these rules once, then run drafts through the matching agent without re-explaining the brief every time. The workflow API handles async batch runs, so editors queue 20 posts Friday afternoon and review flagged issues Monday morning. This works if your style guides are stable enough to configure upfront. If clients change voice quarterly or you're still defining the rules, you'll spend more time updating profiles than the MCP saves. Best fit: agencies with 5+ recurring clients and documented style standards.

Using Markup AI to lock down product vocabulary across writers

A 3-person docs team at a SaaS company ships API references, user guides, and SDK tutorials. The product uses 40+ domain-specific terms—some inherited from legacy systems, some newly coined—and inconsistent usage confuses developers. Markup AI's terminology resources let you build a canonical term set, then run new docs through an agent that flags deviations before merge. The MCP's list and get endpoints mean you can surface term definitions in Switchy without leaving the editor. This pays off when your vocabulary is large enough that writers can't memorize it but stable enough that you're not redefining terms weekly. If you're pre-launch or rebranding, wait until terminology settles. Ideal threshold: 20+ terms, 2+ writers, monthly doc cadence.

When Markup AI handles pre-publish legal checks at small scale

A 4-person marketing team at a fintech startup writes blog posts, email campaigns, and landing pages that must avoid specific claims before legal review. They've documented 15 prohibited phrases and 8 required disclosures in a compliance guide. Markup AI's style guide profiles encode these rules, and the workflow API runs drafts through a compliance agent that flags violations before the legal team sees them. This cuts review cycles from 3 days to 1 because legal only sees clean drafts. The trade-off: setup cost is high if your compliance rules change monthly or span 100+ edge cases. Best for teams with stable, medium-complexity compliance requirements—regulated industries where the rules are clear but manual checking is slow.