SEO Optimized Blog Generator Agent Template

What it does: Coordinates six specialist agents (Orchestrator, Research, Outline, Drafter, Editor, Publisher) to turn a topic plus seed keywords into an SEO-ready B2B blog post delivered as a Google Doc, with two human review checkpoints.
Best for: content marketers, founders, and SEO teams who want a repeatable, company-aware blog pipeline from a one-line request.
Apps used: Web Search (built-in), DataForSEO (optional), Google Docs.
Setup time: 10 to 20 minutes, including a one-time onboarding interview.

SEO Optimized Blog Generator is a no-code, six-agent workflow that turns a topic and seed keywords into a finished, SEO-ready blog post. An Orchestrator coordinates the Research, Outline, Drafter, Editor, and Publisher agents, then delivers the post to Google Docs with a headline and keyword pack. This template acts as a starting point: any app, instructions, and agents can be updated as required for your own use case, simply ask Max (Agent Builder) to update the workflow as needed.

Last verified from workflow config on June 2026. Includes common failure modes and fixes.

Problem this solves

Producing SEO blog posts in-house is slow: keyword research, briefing, drafting, editing, and formatting are separate manual jobs.
Generic AI drafts read as machine-written and skip real keyword and SERP grounding, so they rank poorly and need heavy rewriting.
Company facts, voice, and approved claims live in people's heads, so every writer reinvents them and risks publishing unverified claims.
You want one repeatable pipeline that researches, drafts, edits for E-E-A-T, and returns a publish-ready Google Doc from a short request.

What this agent does

This template can:

Onboard your company once through a short interview and save a reusable company profile (positioning, ICP, voice, banned words, categories, approved claims, CTAs).
Validate topic demand and pull keyword and SERP data with DataForSEO, falling back to web search when DataForSEO is not connected.
Build a content brief: primary and secondary keywords, search intent, target word count, internal and external link plan, and an FAQ list.
Find a sharp angle and differentiation, and gather real practitioner sources for E-E-A-T texture.
Turn the research into a MECE outline, then a full draft that follows the chosen angle and word target.
Edit for conciseness, scannability, E-E-A-T, and answer-engine formatting (tables, bullets, short sentences).
Enforce voice rules and hard bans (no em or en dashes, no AI filler) and reject structurally failed drafts for rework.
Generate headline options, meta title, meta description, and slug, then run a full QA checklist.
Assemble a CMS-ready post (frontmatter plus body) and deliver it as a formatted Google Doc with links, lists, and tables intact.
Pause for two review checkpoints (outline and draft) unless you run it fully unattended.

How it works

First run (onboarding): Send "setup" (or "onboarding"). The Orchestrator runs a short interview, about four turns, and saves your company profile to a shared references file that every future run reuses.
Write a post: Send a short message with topic plus seed keywords, for example "Reducing SaaS churn | keywords: churn rate, retention, B2B onboarding". Optional flags: "fully unattended", "word target: 1500", or a fixed title kept verbatim as the H1.
Research: The Research Agent runs idea and win condition, SERP and content brief, angle and differentiation, and practitioner sourcing, using web search and DataForSEO.
Outline + Checkpoint 1: The Outline Agent turns the research into an executable outline. You approve the outline or list the changes you want.
Draft + Checkpoint 2: The Drafter Agent writes the full draft from the outline and research. You approve the draft for editing or list changes.
Edit and QA: The Editor Agent runs the edit and E-E-A-T pass, headline and meta, then a full QA checklist. It can reject a weak draft and send it back for rework.
Publish: The Publisher Agent assembles the final post and creates a Google Doc, then returns the doc link, final headline, keyword pack, and a short run summary.

The agent team

Agent	Pipeline role	What it produces
Orchestrator	Coordinates the run, onboarding, and checkpoints	Run brief, company profile, routing
Research	Idea, SERP brief, angle, practitioner sourcing	Keyword brief, angle, sourced texture
Outline	Turns research into a blueprint	Approved outline (Checkpoint 1)
Drafter	Writes the full draft	Draft following angle and word target (Checkpoint 2)
Editor	Edit, E-E-A-T, headline, meta, QA	Edited draft, metadata, QA report
Publisher	Assembles and ships the final post	CMS-ready post and Google Doc

Requirements

A SketricGen account with this workflow cloned into AgentSpace.
Google account (recommended): connect Google Docs so the Publisher delivers the final post as a Doc.
DataForSEO account (optional): connect for keyword volume and SERP data; without it, research falls back to web search.
Web Search: already enabled on the Research and Editor agents, no setup needed.
Inputs: your onboarding answers for the first run, then a topic and seed keywords for each post.

Setup guide

Clone the template and open it in AgentSpace.
Connect Google Docs (recommended): Click the Google Docs tool node on the Publisher Agent, then Connect, and link your Google account so the final post is delivered as a Doc.
Connect DataForSEO (optional): Click the DataForSEO node on the Research Agent, click Connect, and enter your DataForSEO login and password. This gives keyword volume and SERP data; research falls back to web search if you skip it.
Web Search: No setup needed; it is already enabled on the Research and Editor agents.
Run onboarding: Send "setup" and answer the short interview. Review and approve the company profile when the Orchestrator presents it.
Write your first post: Send a topic plus seed keywords. Approve the outline and the draft at the two checkpoints, or add "fully unattended" to skip them.
Open the Google Doc to review, edit, or publish.

Common issues and fixes

Pipeline asks for onboarding on every run: the company profile was not saved or is empty. Re-run "setup", complete the interview, and approve the profile so it writes to the shared references file.
No keyword volume or SERP data: DataForSEO is not connected. Connect it on the Research Agent, or accept the web-search fallback for lighter research.
Final Google Doc not created: Google Docs is not connected or lacks create-document permission. Connect the Publisher's Google Docs tool and re-run; the post is still assembled as a file for manual upload.
Draft keeps getting rejected: the Editor found a structural failure (wrong angle, missed search intent, or unverifiable claims). Read the reject notes, tighten the topic or outline, and re-run.
Post reads generic: add more context in your topic message or refine the company profile (voice, approved claims, pillars) so research and drafting have more to work with.

Customization knobs

Update the company profile any time: send "update profile" to change voice, categories, tags, author, CTAs, or approved claims.
Swap or add research tools (for example, a different SEO API) on the Research Agent and update its instructions.
Change the default word count, or set "word target: N" per run.
Skip the human checkpoints with "fully unattended" for hands-off runs.
Replace Google Docs with another destination, for example Notion, by changing the Publisher's tool and save step. Ask Max (Agent Builder) to make the swap.

Apps used

App	What it is used for	Typical permission scope
Web Search	topic demand, competitor coverage, freshness checks, and claim verification on the Research and Editor agents	built-in, no connection needed
DataForSEO	keyword volume, SERP data, and keyword ideas during research (optional)	DataForSEO login and password (keyword and SERP data tools)
Google Docs	create and deliver the final blog post as a formatted Doc	create-document

(Scopes vary by how you connect each app.)

Use cases

Repeatable SEO blog production: Turn a topic into a researched, publish-ready post on a schedule. Best for content and SEO teams.
Founder-led content: Produce on-brand posts without hiring a full content team. Best for early-stage founders.
Company-aware drafts: Keep voice, approved claims, and categories consistent across many writers. Best for marketing leads.
First drafts at scale: Run fully unattended to build a backlog of drafts for human polish. Best for content agencies.
E-E-A-T-focused posts: Get edited, sourced, QA-checked drafts instead of raw AI text. Best for trust-sensitive niches.

Example prompts and outputs

Starter prompts (copy/paste)

Onboarding (first run): setup
Write a post: Reducing SaaS churn | keywords: churn rate, retention, B2B onboarding

Example 1: Default run

Input: Reducing SaaS churn | keywords: churn rate, retention, B2B onboarding

Output: Research runs (idea, brief, angle, sources). You approve the outline, then the draft. The Editor edits and QA-checks it, and the Publisher delivers a Google Doc with the post, final headline, and keyword pack.

Example 2: With flags

Input: Email deliverability for B2B | keywords: spam filters, SPF, DKIM | word target: 1500 | fully unattended

Output: The pipeline runs end to end without checkpoints, holds the post to about 1,500 words, and delivers the Google Doc plus a short run summary.

Example 3: Fixed title

Input: a topic message that includes a required H1. The pipeline keeps that title verbatim as the post's H1 and writes the rest around it.

Why you need this template

One short message turns into a publish-ready blog post, not a week of handoffs.
Real keyword and SERP grounding plus E-E-A-T editing, so drafts need less rewriting.
A saved company profile keeps voice, claims, and taxonomy consistent across every post.
Two review checkpoints keep you in control, or run unattended when you trust the output.
Editable in AgentSpace: swap tools, change word counts, or change the output destination.