Atom AI Labs - AI-Powered Multi-Tenant Platform

ATOM Marketplace User Guide

Complete guide to using the ATOM marketplace for sharing agents, skills, canvases, and domains.

---

Overview
Getting Started
Browsing the Marketplace
Publishing to Marketplace
Installing from Marketplace
Private Marketplace
Pricing and Billing
Code Examples
Troubleshooting

---

Overview

The ATOM marketplace enables you to:

**Discover** pre-built agents, skills, canvases, and domains
**Share** your creations with the community
**Install** ready-to-use components
**Rate** and review marketplace items
**Federate** across ATOM instances

Marketplace Types

Marketplace	Description	Example Items
Agent Marketplace	Mature AI agents with pre-learned capabilities	Customer support agent, Sales assistant
Skill Marketplace	Reusable AI skills and functions	Data extraction, Document summarization
Canvas Marketplace	UI canvas templates and visual components	Dashboard layout, Workflow builder
Domain Marketplace	Specialist domain expertise templates	Finance specialist, Legal analyst

Skills vs Capabilities

**Important Distinction:**

Aspect	Skills	Capabilities
Definition	Basic composable units (building blocks)	Higher-level competencies that combine multiple skills
Type	Individual tools/commands	Unstructured complex tasks (not long-horizon)
Examples	Text summarization, Data extraction, API call	Data analysis, Code execution, Integrations, Reasoning
Governance	Safety-based governance (LOW_RISK/HIGH_RISK levels)	Domain-based governance (graduation thresholds)
Tracking	SkillExecution records	Capability matrix and graduation service
Domains	None (basic units)	4 domains: DATA_ANALYSIS, CODE_EXECUTION, INTEGRATIONS, REASONING

**How They Work Together:**

Skills (Basic Units) → Combined into → Capabilities (Higher-Level Competencies)
    ↓                                          ↓
Individual tools like:                 "Data Analysis" capability includes:
- Text summarization                    - Data visualization
- Sentiment analysis                    - Chart generation
- Data extraction                       - Report generation
                                          Domain: DATA_ANALYSIS
                                          Risk factor: 0.3
                                          Threshold: 70% success rate

**Key Differences:**

**Skills** are the fundamental building blocks - individual functions you can install and use
**Capabilities** are higher-level competencies that agents develop by combining multiple skills
**Skills** have safety-based governance (LOW_RISK/HIGH_RISK) and require admin approval for dangerous operations
**Capabilities** have domain-based governance with graduation thresholds for agent maturity progression
**Skills** are installed from the marketplace; **Capabilities** are developed by agents through training and experience

**Example:**

To become a "Data Analysis Specialist," an agent needs:

**Skills installed**: Data extraction, Chart rendering, Report generation (from marketplace)
**Capabilities developed**: DATA_ANALYSIS domain proficiency (through training and success rate)

---

Getting Started

Accessing the Marketplace

**Navigate to Marketplace**

Log in to your ATOM workspace
Click **"Marketplace"** in the left sidebar
Choose a marketplace type (Agent, Skill, Canvas, Domain)

**Marketplace Dashboard**

Featured items
Popular categories
Recent submissions
Your installations

Permissions

**Browse**: Available to all users
**Install**: Requires appropriate plan limits
**Publish**: Requires account with publishing permissions
**Rate**: Requires installation of the item

---

Browsing the Marketplace

Search and Filter

**Basic Search:**

Search bar: "customer support"

**Advanced Filters:**

**Category**: Browse by category (e.g., "automation", "integration", "analytics")
**Type**: Filter by type (e.g., "api", "workflow", "template")
**Price**: Free vs. Paid items
**Rating**: Minimum rating (4+ stars, 3+ stars, etc.)
**Sort By**: Rating, installs, updated, name

**Example Filter Combinations:**

Category: "integration"
Type: "api"
Price: "free"
Rating: "4+ stars"
Sort By: "rating"

Viewing Item Details

Click any item to view:

**Overview**

Name and description
Publisher information
Version history
Installation count

**Details**

Long description
Screenshots/demo
Input/output schemas
Dependencies
Resource requirements

**Ratings and Reviews**

Average rating
Total reviews
User reviews with comments
Rating distribution

**Installation**

Install button
Version selector
Compatibility check
Resource impact

Publishing to Marketplace

Preparation Checklist

Before publishing, ensure:

[ ] Item is fully functional and tested
[ ] Documentation is complete
[ ] No hardcoded credentials or sensitive data
[ ] Compatible with current platform version
[ ] Meets quality standards
[ ] Proper licensing information

Publishing an Agent

**Requirements:**

Agent maturity: **Autonomous** or **Supervised**
Minimum performance metrics:
Success rate: ≥70%
Zero-intervention ratio: ≥60%
Constitutional compliance: ≥80%
Graduation exam passed (if applicable)

**Steps:**

**Navigate to Your Agents**

**Configure Publishing Details**

**Select Version**

Choose which version to publish
Include agent memory (optional)
Include learned experiences (optional)
Strip sensitive data automatically

**Add Documentation**

Long description (Markdown supported)
Usage instructions
Screenshots or demo video
Changelog

**Submit for Review**

Review submission
Agree to publisher terms
Submit

**Approval Process:**

Automated validation (5-10 minutes)
Manual review (1-2 business days)
Approval notification via email
Item appears in marketplace

Publishing a Skill

**Requirements:**

Skill code is functional
Input/output schemas defined
Error handling implemented
Documentation complete

**Steps:**

**Navigate to Your Skills**

**Configure Skill Details**

**Define Schemas**

**Input Schema:**

```json

{

"type": "object",

"properties": {

"text": {

"type": "string",

"description": "Text to extract data from"

"fields": {

"type": "array",

"items": {"type": "string"},

"description": "Fields to extract"

}

"required": ["text", "fields"]

}

```

**Output Schema:**

```json

{

"type": "object",

"properties": {

"extracted_data": {

"type": "object",

"description": "Extracted key-value pairs"

"confidence": {

"type": "number",

"description": "Confidence score (0-1)"

}

```

**Upload Code**

Paste skill code
Specify language (Python, JavaScript)
Add dependencies

**Add Examples**

**Submit for Review**

Publishing a Canvas

**Requirements:**

Canvas is functional
Components are properly connected
Layout is responsive
Documentation included

**Steps:**

**Navigate to Your Canvases**

**Configure Canvas Details**

**Export Canvas Configuration**

Components included
Layout preserved
Dependencies listed
Data sources specified (placeholders)

**Add Screenshots**

Upload canvas screenshot
Add demo video (optional)
Document component interactions

**Submit for Review**

Publishing a Domain

**Requirements:**

Domain is fully configured
Knowledge base is populated
Skills are integrated
Documentation complete

**Steps:**

**Navigate to Your Domains**

**Configure Domain Details**

**Export Domain Configuration**

Knowledge base (anonymized)
Integrated skills
Agent templates
Best practices

**Submit for Review**

---

Installing from Marketplace

Installation Prerequisites

**Plan Limits**: Check your plan allows additional items
**Resource Requirements**: Ensure sufficient resources
**Compatibility**: Verify platform version compatibility
**Dependencies**: Check required integrations/skills

Installing an Agent

**Steps:**

**Browse Agent Marketplace**

**Select Agent**

Read description
Check ratings and reviews
View resource requirements
Verify compatibility

**Click "Install"**

Choose version (latest or specific)
Review permissions
Confirm installation

**Configure Installed Agent**

**Agent Ready**

Agent appears in your Agents list
Can be used immediately
Can be further trained

**Example:**

Installing: Customer Support Agent v2.1
Publisher: Acme Corp
Rating: ⭐⭐⭐⭐⭐ (4.8/5 from 120 reviews)
Resource Impact: Medium
Estimated Cost: $0/month (included in plan)

✓ Compatible with your plan
✓ No additional integrations required
✓ Sufficient resources available

[Install Agent]

Installing a Skill

**Steps:**

**Browse Skill Marketplace**

**Select Skill**

Review description
Check input/output schemas
View examples
Verify dependencies

**Click "Install"**

Review skill code
Check dependencies
Confirm installation

**Skill Available**

Skill appears in Canvas skill palette
Can be used in workflows
Can be customized

**Example:**

Installing: Data Extractor v1.3
Publisher: DataTools Inc
Rating: ⭐⭐⭐⭐ (4.2/5 from 45 reviews)
Type: API
Dependencies: None

Input Schema: {text: string, fields: array}
Output Schema: {extracted_data: object, confidence: number}

Example:
Input: {"text": "Contact: john@example.com", "fields": ["email"]}
Output: {"extracted_data": {"email": "john@example.com"}, "confidence": 0.95}

[Install Skill]

Installing a Canvas

**Steps:**

**Browse Canvas Marketplace**

**Select Canvas**

View screenshots
Check components
Review documentation
Verify data sources

**Click "Install"**

Review canvas configuration
Map data sources
Confirm installation

**Canvas Ready**

Canvas appears in your Canvas list
Can be customized
Can be shared with team

**Example:**

Installing: Sales Dashboard v1.0
Publisher: DashboardTemplates
Rating: ⭐⭐⭐⭐⭐ (4.9/5 from 89 reviews)
Components: 12
Data Sources: Salesforce, Stripe

⚠️ Requires: Salesforce integration
⚠️ Requires: Stripe integration

[Configure Data Sources] [Install Canvas]

Installing a Domain

**Steps:**

**Browse Domain Marketplace**

**Select Domain**

Read description
Check features
View documentation
Verify pricing

**Click "Install"**

Review pricing
Check compatibility
Confirm installation

**Configure Domain**

Import knowledge base
Configure agents
Set up skills

**Domain Ready**

Domain available for agents
Can be customized
Subscription activated

**Example:**

Installing: Finance Specialist v3.0
Publisher: FinanceAI
Rating: ⭐⭐⭐⭐ (4.5/5 from 56 reviews)
Price: $9.99/month
Subscription: Monthly (cancel anytime)

Features:
✓ Financial analysis
✓ Report generation
✓ Forecasting
✓ Compliance checking

[Subscribe & Install]

Managing Installations

**View Your Installations:**

Marketplace → "My Installations"

**Update Installed Items:**

Check for updates
View changelog
Update to latest version

**Uninstall Items:**

Remove from workspace
Data retention options
Reinstall capability

---

Private Marketplace

Overview

Private marketplace allows organizations to:

Share items within organization only
Control access and approvals
Maintain internal asset library
Separate from public marketplace

Setting Up Private Marketplace

**Enable Private Marketplace:**

Settings → Marketplace → "Enable Private Marketplace"

**Configuration:**

Private Marketplace Name: "Acme Corp Internal"
Visibility: Organization only
Approval Required: Yes/No
Allow Publishing: All users / Admins only

Publishing to Private Marketplace

**Steps:**

**Navigate to Item**

Agent/Skill/Canvas/Domain
Click "Publish"

**Choose Private Marketplace**

**Configure Details**

Name and description
Category and tags
Access control (if applicable)

**Submit**

Instant approval (if enabled)
Or submit for review

Accessing Private Marketplace

**View Private Marketplace:**

Marketplace → "Private" tab

**Filter by:**

Publisher
Department
Date published
Installation count

Private Marketplace Features

**Access Control:**

Restrict by user role
Restrict by department
Require approval for access

**Versioning:**

Multiple versions
Version approvals
Rollback capability

**Analytics:**

Installation tracking
Usage metrics
Feedback collection

---

Pricing and Billing

Marketplace Pricing Models

**Free Items:**

No cost to install
Included in plan limits
Full functionality

**Paid Items:**

One-time purchase or subscription
Billed through ATOM account
Can cancel subscriptions anytime

Plan Limits

Plan	Max Agents	Max Skills	Max Canvases	Max Domains
Free	3	10	5	1
Solo	10	50	20	5
Team	25	200	100	10
Enterprise	Unlimited	Unlimited	Unlimited	Unlimited

Resource Usage

**Agent Resources:**

CPU: Low/Medium/High
Memory: 100MB - 2GB
Storage: Included in plan

**Skill Resources:**

Execution time: Included
API calls: Count toward quota
Storage: Minimal

**Canvas Resources:**

Components: No limit
Data sources: Per integration limits
Storage: Included in plan

**Domain Resources:**

Knowledge base: Per plan storage
Agents: Count toward plan limit
Skills: Count toward plan limit

Billing

**Free Items:**

No additional charges
Count toward plan limits

**Paid Items:**

Billed monthly
Itemized on invoice
Can cancel anytime

**Example Invoice:**

ATOM SaaS - Monthly Invoice
----------------------------
Plan: Team ($79/month)
Marketplace Items:
  - Finance Specialist Domain: $9.99
  - Sales Dashboard Canvas: $4.99
Total: $93.98

---

Rating and Reviews

Rating Items

**Rate After Installation:**

Use the item for reasonable period
Click "Rate" button
Select star rating (1-5)
Write review (optional)
Submit

**Rating Guidelines:**

**5 stars**: Excellent, exceeds expectations
**4 stars**: Good, meets expectations
**3 stars**: Average, some issues
**2 stars**: Poor, significant issues
**1 star**: Unusable, critical issues

Writing Reviews

**What to Include:**

What you like
What could be improved
Use case context
Tips for other users

**Example Review:**

⭐⭐⭐⭐⭐
"Excellent customer support agent! Handles inquiries across email
and chat seamlessly. Reduced response time by 80%. Highly recommend
for small businesses."

Use case: Customer support for e-commerce store
Team size: 5 agents
Integration: Gmail, Slack

Review Moderation

**Community Guidelines:**

Be respectful and constructive
Avoid spam and self-promotion
Report inappropriate reviews
Edit your own reviews

---

Code Examples

API Usage Examples

Browse Marketplace via API

**Python Example:**

import requests

# Browse skills marketplace
response = requests.get("https://api.atomagentos.com/api/public/v1/marketplace/skills")
data = response.json()

for skill in data['skills']:
    print(f"{skill['name']}: {skill['description']}")
    print(f"Rating: {skill['average_rating']}/5")
    print(f"Installs: {skill['install_count']}")
    print()

**JavaScript Example:**

// Browse agents marketplace
async function browseAgents(category = null) {
  const params = new URLSearchParams({
    limit: 20,
    sort_by: 'rating',
    sort_order: 'desc'
  });

  if (category) {
    params.append('category', category);
  }

  const response = await fetch(
    `https://api.atomagentos.com/api/public/v1/marketplace/agents?${params}`
  );

  const data = await response.json();
  return data.agents;
}

// Usage
const agents = await browseAgents('customer-service');
agents.forEach(agent => {
  console.log(`${agent.name}: ${agent.average_rating}/5`);
});

**Go Example:**

package main

import (
    "encoding/json"
    "fmt"
    "net/http"
)

type Skill struct {
    ID          string  `json:"id"`
    Name        string  `json:"name"`
    Description string  `json:"description"`
    Rating      float64 `json:"average_rating"`
}

func main() {
    resp, err := http.Get("https://api.atomagentos.com/api/public/v1/marketplace/skills")
    if err != nil {
        panic(err)
    }
    defer resp.Body.Close()

    var data struct {
        Skills []Skill `json:"skills"`
    }

    json.NewDecoder(resp.Body).Decode(&data)

    for _, skill := range data.Skills {
        fmt.Printf("%s: %.1f/5\n", skill.Name, skill.Rating)
    }
}

**cURL Example:**

# Browse skills with filters
curl -X GET "https://api.atomagentos.com/api/public/v1/marketplace/skills?category=data-processing&type=api&sort_by=rating&sort_order=desc" \
  -H "Accept: application/json"

# Get specific skill details
curl -X GET "https://api.atomagentos.com/api/public/v1/marketplace/skills/SKILL_ID" \
  -H "Accept: application/json"

# Rate a skill
curl -X POST "https://api.atomagentos.com/api/public/v1/marketplace/skills/SKILL_ID/rate" \
  -H "X-API-Key: atom_pub_YOUR_KEY_HERE" \
  -H "Content-Type: application/json" \
  -d '{"rating": 5, "review": "Excellent skill!"}'

Install via API

**Python Example:**

import requests

# Install a skill
api_key = "atom_pub_YOUR_KEY_HERE"
headers = {"X-API-Key": api_key}

response = requests.post(
    f"https://api.atomagentos.com/api/public/v1/marketplace/skills/SKILL_ID/install",
    headers=headers
)

if response.status_code == 200:
    print("Skill installed successfully!")
else:
    print(f"Installation failed: {response.json()['error']}")

**JavaScript Example:**

// Install an agent
async function installAgent(agentId) {
  const response = await fetch(
    `https://api.atomagentos.com/api/public/v1/marketplace/agents/${agentId}/install`,
    {
      method: 'POST',
      headers: {
        'X-API-Key': 'atom_pub_YOUR_KEY_HERE',
        'Content-Type': 'application/json'
      }
    }
  );

  if (response.ok) {
    const data = await response.json();
    console.log('Agent installed:', data.name);
    return data;
  } else {
    throw new Error('Installation failed');
  }
}

// Usage
installAgent('AGENT_ID')
  .then(agent => console.log('Success:', agent))
  .catch(err => console.error('Error:', err));

Publish via API

**Python Example:**

import requests
import json

api_key = "atom_pub_YOUR_KEY_HERE"
headers = {"X-API-Key": api_key, "Content-Type": "application/json"}

# Prepare skill data
skill_data = {
    "name": "Data Extractor Pro",
    "description": "Advanced data extraction from unstructured text",
    "long_description": "# Data Extractor Pro\n\nExtracts structured data from text using NLP.",
    "type": "api",
    "category": "data-processing",
    "tags": ["extraction", "nlp", "parsing"],
    "price": 0.0,
    "license": "MIT",
    "input_schema": {
        "type": "object",
        "properties": {
            "text": {
                "type": "string",
                "description": "Text to extract from"
            },
            "fields": {
                "type": "array",
                "items": {"type": "string"},
                "description": "Fields to extract"
            }
        },
        "required": ["text", "fields"]
    },
    "output_schema": {
        "type": "object",
        "properties": {
            "extracted_data": {
                "type": "object",
                "description": "Extracted key-value pairs"
            },
            "confidence": {
                "type": "number",
                "description": "Confidence score (0-1)"
            }
        }
    },
    "code": """
def execute(input):
    text = input['text']
    fields = input['fields']
    # Extraction logic here
    return {
        'extracted_data': {},
        'confidence': 0.95
    }
"""
}

# Submit skill
response = requests.post(
    "https://api.atomagentos.com/api/public/v1/marketplace/submit",
    headers=headers,
    json=skill_data
)

if response.status_code == 201:
    print("Skill submitted for review!")
    print(f"Submission ID: {response.json()['id']}")
else:
    print(f"Submission failed: {response.json()['error']}")

---

CLI Examples

Using ATOM CLI

**Install ATOM CLI:**

# macOS/Linux
curl -fsSL https://get.atomagentos.com/install.sh | bash

# Windows (PowerShell)
irm https://get.atomagentos.com/install.ps1 | iex

**Browse Marketplace:**

# List all skills
atom marketplace list skills

# Search skills
atom marketplace search skills --query "data extraction"

# Get skill details
atom marketplace get skill SKILL_ID

# List categories
atom marketplace categories

**Install Items:**

# Install a skill
atom marketplace install skill SKILL_ID

# Install specific version
atom marketplace install skill SKILL_ID --version 2.1.0

# Install an agent
atom marketplace install agent AGENT_ID

# Install multiple items
atom marketplace install skill SKILL_1 skill SKILL_2 skill SKILL_3

**Publish Items:**

# Publish a skill from file
atom marketplace publish skill --file skill.json

# Publish an agent
atom marketplace publish agent --agent-id my_agent --price 0.0

# Update published item
atom marketplace update skill SKILL_ID --file skill_v2.json

# Check submission status
atom marketplace status SUBMISSION_ID

**Rate Items:**

# Rate a skill
atom marketplace rate skill SKILL_ID --rating 5 --review "Excellent!"

# List your ratings
atom marketplace ratings

# Get item ratings
atom marketplace ratings skill SKILL_ID

**Manage Installations:**

# List installed items
atom marketplace installed

# Check for updates
atom marketplace updates

# Update an item
atom marketplace update skill SKILL_ID

# Uninstall an item
atom marketplace uninstall skill SKILL_ID

---

Example Configurations

Skill Configuration Example

**Simple API Skill:**

{
  "name": "Text Summarizer",
  "description": "Summarizes long text into concise summaries",
  "type": "api",
  "category": "text-processing",
  "tags": ["summarization", "nlp", "text"],
  "price": 0.0,
  "license": "MIT",
  "input_schema": {
    "type": "object",
    "properties": {
      "text": {
        "type": "string",
        "description": "Text to summarize"
      },
      "max_length": {
        "type": "integer",
        "description": "Maximum summary length",
        "default": 100
      }
    },
    "required": ["text"]
  },
  "output_schema": {
    "type": "object",
    "properties": {
      "summary": {
        "type": "string",
        "description": "Generated summary"
      },
      "original_length": {
        "type": "integer",
        "description": "Original text length"
      },
      "summary_length": {
        "type": "integer",
        "description": "Summary length"
      }
    }
  },
  "code": "def execute(input):\n    text = input['text']\n    max_length = input.get('max_length', 100)\n    # Summarization logic\n    summary = text[:max_length] + '...'\n    return {\n        'summary': summary,\n        'original_length': len(text),\n        'summary_length': len(summary)\n    }"
}

**Integration Skill Example:**

{
  "name": "Slack Notifier",
  "description": "Send notifications to Slack channels",
  "type": "integration",
  "category": "communication",
  "tags": ["slack", "notifications", "messaging"],
  "price": 0.0,
  "license": "Apache-2.0",
  "dependencies": [
    {"name": "slack_sdk", "version": ">=3.0.0"}
  ],
  "input_schema": {
    "type": "object",
    "properties": {
      "channel": {
        "type": "string",
        "description": "Slack channel ID or name"
      },
      "message": {
        "type": "string",
        "description": "Message to send"
      },
      "webhook_url": {
        "type": "string",
        "description": "Slack webhook URL"
      }
    },
    "required": ["channel", "message", "webhook_url"]
  },
  "output_schema": {
    "type": "object",
    "properties": {
      "success": {
        "type": "boolean",
        "description": "True if message sent successfully"
      },
      "timestamp": {
        "type": "string",
        "description": "ISO timestamp of message"
      }
    }
  },
  "code": "import requests\n\ndef execute(input):\n    webhook_url = input['webhook_url']\n    payload = {\n        'channel': input['channel'],\n        'text': input['message']\n    }\n    response = requests.post(webhook_url, json=payload)\n    return {\n        'success': response.status_code == 200,\n        'timestamp': response.headers.get('Date', '')\n    }"
}

Agent Template Example

**Customer Support Agent:**

{
  "name": "Customer Support Agent",
  "description": "Handles customer inquiries across email and chat",
  "category": "customer-service",
  "tags": ["support", "customer-service", "helpdesk"],
  "version": "2.1.0",
  "maturity_level": "supervised",
  "capabilities": [
    "respond_to_inquiries",
    "escalate_issues",
    "lookup_orders",
    "process_returns"
  ],
  "skills": [
    "text_understanding",
    "sentiment_analysis",
    "knowledge_base_search"
  ],
  "integrations": [
    "gmail",
    "slack",
    "salesforce"
  ],
  "memory": {
    "include": true,
    "anonymize": true,
    "retention_days": 90
  },
  "performance_metrics": {
    "success_rate": 0.85,
    "zero_intervention_ratio": 0.72,
    "constitutional_compliance": 0.92
  },
  "configuration": {
    "max_response_time": 300,
    "escalation_threshold": 0.7,
    "learning_enabled": true
  }
}

Canvas Template Example

**Dashboard Canvas:**

{
  "name": "Sales Dashboard",
  "description": "Comprehensive sales analytics dashboard",
  "category": "dashboards",
  "tags": ["sales", "analytics", "dashboard"],
  "version": "1.0.0",
  "layout": {
    "type": "grid",
    "columns": 3,
    "rows": 4
  },
  "components": [
    {
      "id": "revenue_chart",
      "type": "chart",
      "position": {"row": 0, "col": 0, "rowspan": 2, "colspan": 2},
      "data_source": "salesforce",
      "query": "SELECT date, revenue FROM sales ORDER BY date DESC LIMIT 30",
      "chart_type": "line"
    },
    {
      "id": "kpi_cards",
      "type": "kpi",
      "position": {"row": 0, "col": 2, "rowspan": 1, "colspan": 1},
      "metrics": [
        {"label": "Total Revenue", "value": "${revenue_total}"},
        {"label": "New Customers", "value": "${customers_count}"},
        {"label": "Conversion Rate", "value": "${conversion_rate}%"}
      ]
    },
    {
      "id": "recent_orders",
      "type": "table",
      "position": {"row": 2, "col": 0, "rowspan": 2, "colspan": 3},
      "data_source": "salesforce",
      "query": "SELECT order_id, customer, amount, status FROM orders ORDER BY created_at DESC LIMIT 10"
    }
  ],
  "data_sources": [
    {
      "name": "salesforce",
      "type": "salesforce",
      "placeholder": true
    }
  ]
}

---

Screenshot Placeholders

Marketplace Browse Screen

┌─────────────────────────────────────────────────────────────┐
│ ATOM Marketplace                                [Search...]  │
├─────────────────────────────────────────────────────────────┤
│                                                              │
│  [Agents] [Skills] [Canvases] [Domains]  [Private]          │
│                                                              │
│  Filters: [Category ▼] [Type ▼] [Price ▼] [Rating ▼]       │
│                                                              │
│  ┌───────────────────────────────────────────────────────┐  │
│  │ Featured: Customer Support Agent                      │  │
│  │ ⭐⭐⭐⭐⭐ 4.8/5  |  1.2k installs  |  Free            │  │
│  │ Handles customer inquiries across email and chat...    │  │
│  │ [View Details] [Install]                               │  │
│  └───────────────────────────────────────────────────────┘  │
│                                                              │
│  ┌─────────────┐ ┌─────────────┐ ┌─────────────┐            │
│  │ Data        │ │ Email       │ │ Slack       │            │
│  │ Extractor   │ │ Sender      │ │ Notifier    │            │
│  │ ⭐ 4.2      │ │ ⭐ 4.5      │ │ ⭐ 4.7      │            │
│  │ [Install]   │ │ [Install]   │ │ [Install]   │            │
│  └─────────────┘ └─────────────┘ └─────────────┘            │
│                                                              │
└─────────────────────────────────────────────────────────────┘

Item Details Screen

┌─────────────────────────────────────────────────────────────┐
│ Data Extractor v1.3                          [Rate] [Share] │
├─────────────────────────────────────────────────────────────┤
│                                                              │
│  ⭐⭐⭐⭐ 4.2/5  |  45 reviews  |  856 installs               │
│                                                              │
│  Extract structured data from unstructured text using NLP   │
│                                                              │
│  [Install Skill] [Try Demo] [View Code]                     │
│                                                              │
│  Overview                                                    │
│  ─────────────────────────────────────────────────────────  │
│  Type: API                                                   │
│  Category: Data Processing                                   │
│  Publisher: DataTools Inc                                    │
│  Version: 1.3 (Updated 2 days ago)                           │
│  Dependencies: None                                          │
│                                                              │
│  Input/Output Example                                        │
│  ─────────────────────────────────────────────────────────  │
│  Input:                                                      │
│  {                                                           │
│    "text": "Contact: john@example.com, phone: 555-1234",    │
│    "fields": ["email", "phone"]                              │
│  }                                                           │
│                                                              │
│  Output:                                                     │
│  {                                                           │
│    "extracted_data": {                                       │
│      "email": "john@example.com",                            │
│      "phone": "555-1234"                                     │
│    },                                                        │
│    "confidence": 0.95                                        │
│  }                                                           │
│                                                              │
│  Reviews (45)                                                │
│  ─────────────────────────────────────────────────────────  │
│  ⭐⭐⭐⭐⭐ "Excellent! Saved me hours of manual work."       │
│  - Jane D., 2 days ago                                       │
│                                                              │
│  ⭐⭐⭐⭐ "Great accuracy, easy to use."                      │
│  - Mike S., 1 week ago                                       │
│                                                              │
└─────────────────────────────────────────────────────────────┘

Installation Screen

┌─────────────────────────────────────────────────────────────┐
│ Install Customer Support Agent                              │
├─────────────────────────────────────────────────────────────┤
│                                                              │
│  Review Installation Details                                 │
│  ─────────────────────────────────────────────────────────  │
│                                                              │
│  Agent: Customer Support Agent v2.1                          │
│  Publisher: Acme Corp                                        │
│  Rating: ⭐⭐⭐⭐⭐ (4.8/5 from 120 reviews)                 │
│                                                              │
│  Resource Impact: Medium                                     │
│  - CPU: ~20% per active conversation                        │
│  - Memory: ~500MB                                            │
│  - Storage: Included in plan                                 │
│                                                              │
│  Capabilities:                                               │
│  ✓ Respond to customer inquiries                             │
│  ✓ Escalate complex issues                                   │
│  ✓ Lookup order information                                  │
│  ✓ Process returns and refunds                               │
│                                                              │
│  Integrations Required:                                      │
│  ✓ Gmail (already connected)                                 │
│  ✓ Slack (needs connection)                                  │
│  ⚠ Salesforce (optional)                                     │
│                                                              │
│  Plan Check:                                                 │
│  ✓ Free plan allows 3 agents (1 used, 2 available)          │
│  ✓ Sufficient resources available                            │
│                                                              │
│  Configuration:                                              │
│  Name: [Customer Support Agent_________________]             │
│  Initial Maturity: [Intern ▼]                                │
│                                                              │
│  [Cancel]                                    [Install Agent] │
│                                                              │
└─────────────────────────────────────────────────────────────┘

Publishing Screen

┌─────────────────────────────────────────────────────────────┐
│ Publish to Marketplace: Data Extractor                       │
├─────────────────────────────────────────────────────────────┤
│                                                              │
│  1. Basic Information                                        │
│  ─────────────────────────────────────────────────────────  │
│  Name: [Data Extractor_______________________]               │
│  Description: [Extract structured data from text____]        │
│  Category: [Data Processing ▼]                               │
│  Tags: [extraction, nlp, parsing__________]                  │
│                                                              │
│  2. Pricing                                                  │
│  ─────────────────────────────────────────────────────────  │
│  Price: [Free ☑]  [Paid $______]                             │
│  License: [MIT ▼]                                            │
│                                                              │
│  3. Documentation                                           │
│  ─────────────────────────────────────────────────────────  │
│  Long Description:                                           │
│  ┌────────────────────────────────────────────────────┐     │
│  │ # Data Extractor                                    │     │
│  │                                                     │     │
│  │ Extracts structured data from unstructured text     │     │
│  │ using advanced NLP techniques.                      │     │
│  │                                                     │     │
│  │ ## Features                                         │     │
│  │ - Email extraction                                  │     │
│  │ - Phone number extraction                           │     │
│  │ - Custom field extraction                           │     │
│  └────────────────────────────────────────────────────┘     │
│                                                              │
│  4. Preview                                                 │
│  ─────────────────────────────────────────────────────────  │
│  Input Schema: ✓ Defined                                     │
│  Output Schema: ✓ Defined                                    │
│  Code: ✓ Valid Python                                        │
│  Examples: ✓ Added                                           │
│                                                              │
│  [Save Draft]           [Cancel]           [Submit for Review]│
│                                                              │
└─────────────────────────────────────────────────────────────┘

---

Integration Examples

Webhook Integration

**Receive Marketplace Updates:**

from flask import Flask, request, jsonify

app = Flask(__name__)

@app.route('/webhook/marketplace', methods=['POST'])
def marketplace_webhook():
    data = request.json

    event_type = data.get('event_type')
    item_id = data.get('item_id')
    item_name = data.get('item_name')

    if event_type == 'item.installed':
        print(f"New installation of {item_name}!")
        # Send notification, update metrics, etc.

    elif event_type == 'item.rated':
        rating = data.get('rating')
        print(f"{item_name} received {rating} stars!")
        # Process rating

    elif event_type == 'item.updated':
        version = data.get('version')
        print(f"{item_name} updated to v{version}")
        # Handle update

    return jsonify({'status': 'received'}), 200

if __name__ == '__main__':
    app.run(port=5000)

Marketplace Analytics

**Track Installation Performance:**

import requests
import matplotlib.pyplot as plt

api_key = "atom_pub_YOUR_KEY_HERE"
headers = {"X-API-Key": api_key}

# Get your published items
response = requests.get(
    "https://api.atomagentos.com/api/public/v1/marketplace/my-items",
    headers=headers
)

items = response.json()['items']

# Analyze performance
for item in items:
    name = item['name']
    installs = item['install_count']
    rating = item['average_rating']
    reviews = item['review_count']

    print(f"{name}:")
    print(f"  Installs: {installs}")
    print(f"  Rating: {rating}/5 ({reviews} reviews)")
    print()

    # Optional: Create visualization
    plt.bar(name, installs)

plt.show()

---

Troubleshooting

Installation Issues

**"Installation Failed" Error:**

Check plan limits
Verify compatibility
Ensure sufficient resources
Check required integrations

**"Already Installed" Message:**

Check your installations
Update existing version
Uninstall before reinstalling

**"Insufficient Resources" Warning:**

Upgrade plan
Remove unused items
Contact support

Publishing Issues

**"Submission Rejected" Message:**

Review rejection reason
Fix identified issues
Resubmit

**"Pending Approval" Status:**

Wait for review (1-2 business days)
Check email for updates
Contact support if delayed

**"Validation Failed" Error:**

Check item requirements
Fix validation errors
Ensure documentation complete

Performance Issues

**Item Slow to Load:**

Check internet connection
Clear browser cache
Try again later
Report if persistent

**Item Not Working:**

Check compatibility
Review documentation
Update to latest version
Contact publisher

Billing Issues

**Unexpected Charges:**

Review invoice details
Check subscription items
Cancel unwanted subscriptions
Contact billing support

**Plan Upgrade Needed:**

Review plan limits
Compare plans
Upgrade plan
Install item after upgrade

Getting Help

**Support Resources:**

**Documentation**: docs.atomagentos.com
**Community**: community.atomagentos.com
**Support**: support@atomagentos.com
**Status**: status.atomagentos.com

**Contact Support:**

Describe the issue
Include steps to reproduce
Attach screenshots (if applicable)
Provide item name and version

---

Best Practices

For Publishers

**Quality:**

Test thoroughly before publishing
Include comprehensive documentation
Provide clear examples
Respond to reviews and feedback

**Updates:**

Maintain items regularly
Document changes in changelog
Test updates before publishing
Communicate with users

**Pricing:**

Price competitively
Offer free trials for complex items
Provide value for cost
Consider subscription vs. one-time

For Users

**Evaluation:**

Read reviews before installing
Check ratings and installation count
Review documentation
Test with free items first

**Installation:**

Start with free items
Install items gradually
Monitor resource usage
Uninstall unused items

**Feedback:**

Rate items after use
Provide constructive reviews
Report issues to publishers
Suggest improvements

---

Team Collaboration

ATOM provides comprehensive team collaboration features that enable multiple users to work together seamlessly within a shared workspace.

Team Members

**Roles and Permissions:**

Role	Permissions	Description
Owner	Full access, billing management, team management	Primary account holder with complete control
Admin	Manage team, install items, configure settings	Can invite/remove members and manage resources
Member	Use agents, skills, canvases	Standard user access
Viewer	Read-only access	Can view but not execute or modify

Inviting Team Members

**Send Invitation:**

**Navigate to Team Settings**

**Click "Invite Member"**

Enter email address
Select role (Admin, Member, Viewer)
Add optional message
Click "Send Invitation"

**Invitation Email Sent**

Recipient receives email with signup link
Link pre-fills their information
They create account and join team automatically

**Example Invitation:**

To: sarah@company.com
Subject: You're invited to join ACME Corp on ATOM

You've been invited to join ACME Corp's workspace on ATOM.
Click below to accept and get started:

[Accept Invitation]

You'll have Member access with full permissions to use agents,
skills, and collaborate with your team.

Managing Invitations

**View Pending Invitations:**

Go to **Settings → Team → Invitations**
See all pending invitations with:

Email address
Role assigned
Date sent
Status (pending, accepted, expired)
Expiration date (7 days from sending)

**Resend Invitation:**

If an invitation expires or the recipient didn't receive it:

Find invitation in **Settings → Team → Invitations**
Click **"Resend"** button
Confirmation message: "Invitation resent to sarah@company.com"
New expiration date: 7 days from resend

**Cancel Invitation:**

Find invitation in pending list
Click **"Cancel"**
Confirm cancellation
Invitation link becomes invalid

Team Member Management

**View Team Members:**

Settings → Team → Members

Shows all active members with:

Name and email
Role
Date joined
Last active
Status (active, inactive)

**Change Member Role:**

Click member's row
Select new role from dropdown
Click "Save Changes"
Member receives notification of role change

**Remove Team Member:**

Click "Remove" next to member
Confirm removal
Member loses access immediately
Their work is preserved (agents, episodes, etc.)

Team Features

**Shared Resources:**

**Agents**: All team members can use shared agents
**Skills**: Installed skills available to entire team
**Canvases**: Collaborative canvas building and sharing
**Episodes**: Shared episodic memory and learning
**Integrations**: OAuth connections managed at team level

**Real-Time Collaboration:**

See active team members
Collaborate on canvases simultaneously
Share agent executions and results
Comment and discuss within the platform

**Activity Tracking:**

Team activity feed shows recent actions
See who executed which agent
Track skill installations and usage
Monitor resource consumption by team

Team Settings

**Workspace Configuration:**

Settings → Team → Settings

Available settings:

**Workspace Name**: Your team/organization name
**Default Role**: Role for new invited members
**Notification Preferences**: Email notifications settings
**Resource Limits**: Per-member quotas (if applicable)

Team Best Practices

**Onboarding New Members:**

Send invitation with clear role expectations
Share relevant documentation
Provide training on key features
Assign a buddy for first-week questions

**Managing Large Teams:**

Use roles appropriately (limit Admins)
Set up team guidelines and policies
Regularly review member access
Monitor resource usage and costs

**Security:**

Remove former employees promptly
Review active sessions regularly
Use appropriate role assignments
Enable two-factor authentication (coming soon)

**Collaboration Tips:**

Share successful agent configurations
Document team-specific workflows
Use consistent naming conventions
Hold regular knowledge-sharing sessions

Troubleshooting Team Issues

**Invitation Not Received:**

Check spam folder
Verify email address is correct
Resend invitation from settings
Contact support if issues persist

**Member Can't Access Resources:**

Verify their role has appropriate permissions
Check if they're using correct account
Ensure they've accepted invitation
Try removing and re-inviting if needed

**Need to Remove Inactive Members:**

Review last active date in member list
Remove members who haven't logged in 90+ days
This frees up seats for your plan limit

**Exceeded Team Seat Limit:**

Review plan limits in billing
Remove inactive members
Upgrade plan if needed
Contact sales for enterprise teams

---

FAQ

**Q: Can I publish items for free?**

A: Yes! You can publish items for free or set a price.

**Q: Do I own items I install?**

A: You have a license to use installed items. See specific item license for details.

**Q: Can I customize installed items?**

A: Yes, installed items can be customized. Updates from publisher may override customizations.

**Q: How do I get support for an item?**

A: Contact the publisher for item-specific issues. Contact ATOM support for platform issues.

**Q: Can I share installed items with my team?**

A: Yes, items are shared across your organization/workspace.

**Q: What happens if I cancel a paid subscription?**

A: You can use the item until the end of the billing period, then it becomes unavailable.

**Q: Can I get a refund?**

A: Contact support for refund requests. Refunds evaluated case-by-case.

**Q: How are items reviewed?**

A: Items undergo automated validation and manual review for quality, security, and compliance.

---

Glossary

**Agent**: AI assistant with learned capabilities and maturity levels (student, intern, supervised, autonomous)
**Skill**: Basic composable unit - an individual tool or function that can be installed from the marketplace. Skills have safety levels (LOW_RISK, HIGH_RISK) and require admin approval for dangerous operations. Examples: Text summarization, Data extraction, API calls.
**Capability**: Higher-level competency that combines multiple skills. Capabilities are organized into 4 domains (DATA_ANALYSIS, CODE_EXECUTION, INTEGRATIONS, REASONING) with domain-specific graduation thresholds and risk factors. Agents develop capabilities through training and experience.
**Canvas**: Visual workflow builder
**Domain**: Specialized knowledge area or capability domain (DATA_ANALYSIS, CODE_EXECUTION, INTEGRATIONS, REASONING)
**Installation**: Adding a marketplace item to your workspace
**Publishing**: Submitting an item to the marketplace
**Rating**: User feedback score (1-5 stars)
**Review**: Written feedback about an item
**Subscription**: Recurring payment for paid items
**Tenant**: Your organization/workspace in ATOM

---

Next Steps

**Browse Marketplace**: Explore available items
**Install First Item**: Get started with a free item
**Publish Your Creation**: Share with the community
**Join Community**: Connect with other users

---

**← Back to Documentation Index**

ATOM Marketplace User Guide

Table of Contents

Overview

Marketplace Types

Skills vs Capabilities

Getting Started

Accessing the Marketplace

Permissions

Browsing the Marketplace

Search and Filter

Viewing Item Details

Categories

Publishing to Marketplace

Preparation Checklist

Publishing an Agent

Publishing a Skill

Publishing a Canvas

Publishing a Domain

Installing from Marketplace

Installation Prerequisites

Installing an Agent

Installing a Skill

Installing a Canvas

Installing a Domain

Managing Installations

Private Marketplace

Overview

Setting Up Private Marketplace

Publishing to Private Marketplace

Accessing Private Marketplace

Private Marketplace Features

Pricing and Billing

Marketplace Pricing Models

Plan Limits

Resource Usage

Billing

Rating and Reviews

Rating Items

Writing Reviews

Review Moderation

Code Examples

API Usage Examples

Browse Marketplace via API

Install via API

Publish via API

CLI Examples

Using ATOM CLI

Example Configurations

Skill Configuration Example

Agent Template Example

Canvas Template Example

Screenshot Placeholders

Marketplace Browse Screen

Item Details Screen

Installation Screen

Publishing Screen

Integration Examples

Webhook Integration

Marketplace Analytics

Troubleshooting

Installation Issues

Publishing Issues

Performance Issues

Billing Issues

Getting Help

Best Practices

For Publishers

For Users

Team Collaboration

Team Members

Inviting Team Members

Managing Invitations

Team Member Management

Team Features

Team Settings

Team Best Practices

Troubleshooting Team Issues

FAQ

Glossary

Next Steps