Team input & AI output
Nous Research · Multi-agent orchestration · Self-improving skills
Reads campaign briefs, proposes strategy, breaks goals into tasks, creates Chimedeck boards with deadlines and dependencies.
Runs weekly SEMrush audits, identifies keyword gaps vs competitors, creates GitHub PRs for auto-fixable issues, pings team for complex ones.
Writes blog posts, social copy, email snippets, all adapted per platform. Repurposes one article into multi-channel content automatically.
Publishes approved content via Playwright (real browser) to social platforms and via MCPs to Webflow, Shopify. Schedules at optimal times.
Monitors Chimedeck board, posts daily progress to Discord, alerts on blocked or delayed tasks, escalates to humans when decisions are needed. Generates weekly reports to Outline Wiki.
Kanban board where every agent task is visible, trackable, and assignable. Team sees exactly what AI is working on and what's next.
Stores brand knowledge, agent skills, content archive, SEO reports, and campaign history. Agent reads and writes here to stay context-aware.
Team communicates with AI agents via channels and @mentions. All approvals, notifications, and ad-hoc commands flow through Discord.
Agent drafts long-form content and reports directly in Google Docs for easy human review, comments, and collaboration before publishing.
SEO issues that require code changes become pull requests, automatically. Team reviews and merges; agent handles the write-up and branch.
Weekly site audits, keyword gap analysis vs competitors, rank tracking for target keywords, and content opportunity discovery, all automated.
Posts to Facebook, Instagram, and LinkedIn using a real Chrome profile, not blocked by platform API restrictions. Behaves like a real human session.
Crawl errors · Core Web Vitals · Backlinks · Schema · Redirects
Agent classifies each issue: auto-fixable vs needs human
Agent writes the fix, opens PR with description
With full context + recommended fix
5-min review instead of 2-week dev queue
Progress tracked week-over-week · Next audit compares improvement
1 keyword or topic, target audience, campaign goal
Adapts tone, format & length per platform
Webflow · Shopify
Playwright → Page
Playwright → Company
Playwright + hashtags
Google Docs → review
Outline Wiki
Why Playwright + Chrome Profile?
Page posts · Images · Links · Scheduled
Feed posts · Captions · Hashtags · Images
Company page · Articles · Updates
Content Creator drafts post (per-platform tone)
PM queues in for review
Team approves (or edits) in Discord
Executor posts at optimal time via Playwright
Confirmation posted back to Discord
"Q3 campaign — 30% traffic growth target, 2-month window, focus on enterprise segment"
Content pillars, channel mix, weekly content cadence, SEO targets, KPIs — full plan posted to Discord for team review
Review plan in thread, request adjustments. Agent revises until sign-off.
Each deliverable becomes a card: assigned (human or agent), deadline, dependencies, status visible to whole team
Content, SEO, social all run in parallel. PM agent posts daily progress to Discord, escalates blocked items.
Sent to Discord · Saved to Outline Wiki · Compared against KPIs
Agent finishes a blog, social post, SEO report, or campaign plan
Procedure, prompt, and result saved as a reusable skill
Automatically. No retraining, no prompt engineering needed
"The only AI marketing team that genuinely improves from your feedback, without retraining or re-prompting."
Every AI decision logged in Outline Wiki — reviewable, searchable, exportable.
Team can pause, redirect, or cancel any agent task at any time via Discord.
Approval thresholds are configurable — tighten or loosen as trust is established.
saved every week
full-time marketing hire equivalent
gets better every month, not worse
