Agents Overview

AICMOHQ runs a multi-agent pipeline designed specifically for solo founders. Instead of a generic chatbot that produces unfocused copy, AICMOHQ separates marketing into six specialized agents that operate as an asynchronous assembly line.

Signals (GitHub / RSS / Web / AI engines)
           │
           ▼
    ┌──────────────┐
    │  Ved (SEO &  │ ──> Audits visibility, discovers pain points and threads
    │  GEO Agent)  │
    └──────────────┘
           │
           ▼
    ┌──────────────┐
    │ Agni (Ads)   │ ──> Scores audience fit, drafts paid variants
    └──────────────┘
           │
           ▼
    ┌──────────────┐
    │Kavi (Creator)│ ──> Drafts platform-native threads, updates, and articles
    └──────────────┘
           │
           ▼
    ┌──────────────┐
    │Chanakya(GTM) │ ──> Enforces 280-char limits, schedules queue, verifies auth
    └──────────────┘
           │
           ├───────────────────────────────┐
           ▼                               ▼
    ┌──────────────┐                ┌──────────────┐
    │Mitra (Social │                │Shravan (Social│
    │ Media Agent) │                │  Listening)  │
    │(reply drafts)│                │(mentions &   │
    └──────────────┘                │ AI citations)│
                                    └──────────────┘

The Assembly Line & Handoff Contracts

Every stage in the pipeline passes typed data to the next stage through verified schemas.

1. Research & Visibility: Ved (AI SEO & GEO Agent)

Input: Product positioning brief, tracked keywords, GitHub commit feeds, subreddits.
Output: SEO/GEO audit scores, AI-citation reports, and ranked Opportunity objects scored by buyer intent, topical relevance, and recency.
Hard Invariant: Ved is read-only. It scans and monitors discussions but never publishes or mutates external state.

2. Paid Fit & Creative: Agni (AI Ads Agent)

Input: Discovered opportunities, drafts, and your configured Ideal Customer Profile.
Output: Audience-fit scores, ad copy and creative variants, and budget/placement recommendations. Weak-fit items are flagged with concrete fixes.
Purpose: Ensures paid spend and organic drafts both target the buyers you actually want.

3. Copy Production: Kavi (AI Creator Agent)

Input: High-scoring opportunities paired with your brand voice guidelines.
Output: Platform-native copy tailored for X (--- thread format), LinkedIn longform, Reddit contextual contributions, or educational blog posts.
Rules: Enforces genuine founder voice without generic marketing buzzwords.

4. Scheduling & Dispatch: Chanakya (AI GTM Agent)

Input: Staged content drafts and desired publication windows.
Output: Enqueued items in content_queue or direct delivery via channel OAuth APIs.
Safety Gate: Evaluates getEffectiveMode(). In Approve mode, items stay in draft state until you sign off in the dashboard or CLI.

Input: Incoming replies, mentions, and quotes on published content.
Output: Contextual reply suggestions staged in your Agents Feed for quick review.

Input: Brand and competitor mentions, sentiment, AI-engine citations, plus traffic sessions and conversions from Google Analytics 4.
Output: Weekly listening digests, sentiment-shift alerts, AI-citation trends, and recommended channel focus for the upcoming cycle.

Operating Modes: Human Control vs Full Autonomy

The entire agent fleet is governed by the Human Approval Invariant.

Mode	How Drafting Works	How Publishing Works	Best For
Approve (Default)	Agents draft content automatically to your queue	Posts wait for your manual confirmation before going live	New workspaces, high-compliance teams
Autonomous	Agents draft content automatically	Scheduled posts publish directly at optimal times	Established brands with dialed positioning

You can set global defaults or apply granular overrides per agent (for example, keeping Chanakya on approve while allowing Ved and Kavi to run freely). Read the full Operating Modes Guide for command-line and dashboard configuration.

Fleet Management from the Terminal

The entire agent workflow is accessible through the aicmo CLI:

# Run Ved across configured communities
aicmo scout

# Draft marketing updates from your recent git commits
aicmo write --from-git

# Review and approve staged items in the content queue
aicmo queue
aicmo queue approve <content-id>

# Run the complete end-to-end loop: git -> scout -> writer -> queue
aicmo ship