Artifact

​

Introduction

​

Brief Overview of the Artifact Toolkit

• Store Temporary Data: Save intermediate results, generated content, logs, and data retrieved during agent operations
• Manage Context: Preserve and retrieve context across multiple agent interactions or workflow steps
• Share Data Between Agents: Enable collaboration by allowing different agents to access and modify files in shared buckets
• Process Multiple File Formats: Read from various document types (PDF, Excel, Word, images) and create text-based files
• Handle Large Outputs: Store data that exceeds LLM context windows or requires persistent access

​

System Integration with ELITEA

​

Step 1: Create Artifact Toolkit

​

Method 1: Create from Toolkits Menu

1. Navigate to Toolkits Menu: Open the sidebar and select Toolkits.
2. Create New Toolkit: Click the + Create button.
3. Select Artifact: Choose Artifact from the list of available toolkit types.
4. Configure Toolkit Details:
   • Toolkit Name: Enter a descriptive name for the toolkit (e.g., “Report Storage”, “Data Processing Artifacts”)
   • Description: Add an optional description explaining the toolkit’s purpose
   • PgVector Configuration: Select a PgVector connection for vector database integration and semantic search (optional)
   • Embedding Model: Select an embedding model for text processing and semantic search capabilities (optional)
   • Bucket Name: Specify the bucket (storage container) name
     • Creating a New Bucket: Enter a new bucket name, and it will be automatically created when first used
     • Using an Existing Bucket: Enter an existing bucket name to connect to it
     • Naming Convention: Use lowercase letters, numbers, and hyphens only (underscores are automatically converted to hyphens)
     • Best Practice: Use descriptive, unique names to avoid conflicts between different agents or workflows
5. Select Tools: In the “Tools” section, select the checkboxes next to the specific tools you want to enable. Enable only the tools your agents will actually use to follow the principle of least privilege.
6. Save Toolkit: Click Save to create the toolkit.

​

Available Tools:

​

Method 2: Create from Canvas (During Conversation)

1. Navigate to Chat: Open the sidebar and select Chat.
2. Start or Open Conversation: Create a new conversation or open an existing one.
3. Access Toolkit Selection: In the PARTICIPANTS section, locate Toolkits and click the + Add toolkit icon.
4. Create New Toolkit: In the “New Toolkit” dialog, click + Create new Toolkit.
5. Choose Toolkit Type: The canvas interface displays available toolkit categories. Navigate to the appropriate category and select Artifact.
6. Configure Settings: Follow the same configuration steps as Method 1 (display name, bucket name, tool selection).
7. Save and Use: Click Save to create the toolkit. It will be immediately available in your current conversation.

​

Method 3: Automatic Creation via Attachment Settings

1. Enable Attachments: When configuring an agent, pipeline, or conversation, toggle the Allow attachments option.
2. Configure Attachment Storage: In the Attachment settings popup:
   • Option A: Select an existing Artifact Toolkit from the dropdown
   • Option B: Select “Create new” and enter a bucket name to automatically create a new Artifact Toolkit
3. Automatic Integration: The toolkit is created and linked to handle file uploads. Files attached in conversations are automatically stored in the specified bucket.

​

Step 2: Use Toolkit in Agents, Pipelines, or Chat

​

In Agents:

1. Navigate to Agents: Open the sidebar and select Agents.
2. Create or Edit Agent: Click + Create for a new agent or select an existing agent to edit.
3. Add Artifact Toolkit:
   • In the “TOOLKITS” section of the agent configuration, click the “+Toolkit” icon
   • Select your Artifact toolkit from the dropdown list
   • The toolkit will be added to your agent with the previously configured tools enabled

​

In Pipelines:

1. Navigate to Pipelines: Open the sidebar and select Pipelines.
2. Create or Edit Pipeline: Either create a new pipeline or select an existing pipeline to edit.
3. Add Artifact Toolkit:
   • In the “TOOLKITS” section of the pipeline configuration, click the “+Toolkit” icon
   • Select your Artifact toolkit from the dropdown list
   • The toolkit will be added to your pipeline with the previously configured tools enabled

​

In Chat:

1. Navigate to Chat: Open the sidebar and select Chat.
2. Start New Conversation: Click +Create or open an existing conversation.
3. Add Toolkit to Conversation:
   • In the chat Participants section, look for the Toolkits element
   • Click the “Add Tools” icon to open the tools selection dropdown
   • Select your configured Artifact toolkit from the list
   • The toolkit will be added to the conversation

User: "Create a file called meeting_notes.txt with today's agenda"
Agent: [Uses createFile tool to store the agenda in the artifact bucket]

User: "Now append the action items we discussed"
Agent: [Uses appendData tool to add new content to meeting_notes.txt]

User: "Show me all files we've created today"
Agent: [Uses listFiles tool to display all files in the bucket]

​

File Types and Capabilities

​

File Types Supported for Creation

​

Unsupported File Types for Creation

​

Security and Access Control

​

Bucket Accessibility and Security

• Project-Specific Buckets: Buckets created using the Artifact Toolkit are specific to the ELITEA project in which they are created. Buckets and files within one project cannot be directly accessed from another separate ELITEA project.
• Agent Access: Within a project, all agents configured with the Artifact Toolkit can potentially access files within their configured buckets. Agents can only access buckets specified in their toolkit configuration.
• User Access via Artifacts Page: All project members with access to the Artifacts page can view, download, and delete files within project buckets, regardless of which agent created them.
• Shared Bucket Usage: If multiple agents or users within a project configure Artifact Toolkits to use the same bucket name, they will share the same storage space. Recommendations:
  • For independent tasks: Use unique bucket names per agent to avoid unintended data interference
  • For collaborative workflows: Use shared buckets intentionally to enable data exchange
  • File naming conventions: Use clear, descriptive filenames with identifiers to prevent accidental overwrites
  • Regular cleanup: Remove obsolete files to maintain organization
• Security Considerations:
  • Users with agent configuration access can potentially access and manipulate bucket data
  • Do not store highly sensitive data in artifacts without additional security measures
  • Consider encryption for sensitive information before storing
  • Be aware of project-level access controls
  • Regularly review and audit artifact usage

​

Artifact Retention and Lifecycle

• Default Retention: Files created in Artifact buckets have a default retention period of 30 days. After this period, files are automatically deleted and cannot be recovered.
• Custom Retention: You can set custom retention periods when creating buckets:
  Using createNewBucket tool:
  - expiration_measure: "days", "weeks", "months", or "years"
  - expiration_value: numeric value (e.g., 7, 30, 90)
  
• Retention Management: View and modify retention settings through the Artifacts page:
  • View current retention settings per bucket
  • Modify retention periods for existing buckets
  • Monitor days remaining until file expiration
• Manual Deletion: Files and buckets can be manually deleted at any time:
  • By agents using the deleteFile tool
  • By users through the Artifacts page
  • Bulk deletion by removing entire buckets

• Project Limit: Each project has a total storage limit of 9 GB for all artifacts combined
• Individual Files: No strict per-file size limit, but best practice is to keep files reasonably sized
• Performance Impact: Very large files may impact agent processing speed
• Monitoring: Check storage usage in the Artifacts page to avoid exceeding limits

Monitor your project’s artifact storage usage regularly. When approaching the 9 GB limit:

• Delete obsolete files
• Download and remove archived data
• Review retention policies
• Consider using external storage for large datasets

​

Real-World Usage Examples

Agent Instruction: "Create a file named 'detailed_report.txt' to store the generated report. 
Append each section incrementally as it's generated."

User: "Analyze all our Q4 sales data and create a comprehensive report"

Agent: "I'll analyze the Q4 sales data and create a detailed report."
[Agent uses createFile tool to create "Q4_Sales_Report.txt"]
[Agent appends sections incrementally using appendData tool]

Agent: "I've created a comprehensive Q4 sales report with the following sections:
- Executive Summary
- Sales by Region
- Top Performing Products
- Year-over-Year Comparison
- Recommendations

The report has been saved to 'Q4_Sales_Report.txt' in your artifacts."

User: "Can you add a section about customer demographics?"

Agent: [Uses appendData tool to add new section]
"I've added the customer demographics section to the report."

Agent Instruction: "Create 'workflow_state.json' to store the initial state. 
Read and update this state in subsequent steps. 
Delete when workflow completes."

Agent 1: "Save processed data to 'shared_data.json' for the analysis agent."
Agent 2: "Retrieve data from 'shared_data.json' to perform analysis."

User: "Gather all customer feedback from last month"

Agent 1: [Uses createFile to save "customer_feedback_march.json"]
"I've collected and stored 247 customer feedback entries in 'customer_feedback_march.json'."

User: "Analyze the sentiment of last month's customer feedback"

Agent 2: [Uses readFile to access "customer_feedback_march.json"]
[Analyzes data and creates "sentiment_analysis_march.txt"]
"Based on 247 feedback entries:
- Positive: 68%
- Neutral: 22%
- Negative: 10%

Full analysis saved to 'sentiment_analysis_march.txt'."

Agent Instruction: "Save copied text as 'clipboard_item_1.txt'. 
Use 'listFiles' to show available items. 
Retrieve specific items by filename."

Agent Instruction: "Create 'agent_log.txt' at execution start. 
Append timestamped log entries throughout operation."

User: "Start processing the customer orders"

Agent: [Creates "order_processing_log.txt" using createFile]
"Started processing. Created log file."

User: "Process order #1234"

Agent: [Processes order, uses appendData to add log entry]
"Order #1234 processed successfully. Updated log."

User: "Process order #1235"

Agent: [Processes order, uses appendData to add log entry]
"Order #1235 processed successfully. Updated log."

User: "Show me the processing log"

Agent: [Uses readFile to retrieve "order_processing_log.txt"]
"Here's the complete processing log:

2024-01-09 10:15:23 - Processing started
2024-01-09 10:16:45 - Order #1234: Processed successfully
2024-01-09 10:18:12 - Order #1235: Processed successfully"

Agent Instruction: "Use 'listFiles' to see available data files. 
Read 'data_file_1.txt' for processing."

User uploads document → Agent reads it → Agent extracts key information → 
Agent creates structured output file

User: [Attaches contract.pdf] "Extract key terms from this contract"

Agent: [Reads the attached PDF using readFile with attachment integration]
[Creates "contract_key_terms.txt" using createFile]
"I've analyzed the contract and extracted the key terms:

- Contract Duration: 24 months
- Payment Terms: Net 30
- Renewal Clause: Automatic with 60-day notice
- Termination: 90-day written notice required

Full extraction saved to 'contract_key_terms.txt' for reference."

User: "Add a summary of the liability clauses"

Agent: [Uses appendData to add to existing file]
"Added liability clause summary to the document."

User: "Create a configuration file for the production environment"

Agent: [Uses createFile to create "prod_config.json"]
"Created production configuration with default settings."

User: "Update the API timeout to 60 seconds"

Agent: [Uses readFile to get current config]
[Uses overwriteData to update the entire file with new timeout]
"Updated API timeout in production configuration."

User: "Show me all configuration files"

Agent: [Uses listFiles tool]
"Available configuration files:
- prod_config.json
- staging_config.json
- dev_config.json"

Agent uses Playwright MCP toolkit to execute UI tests → 
Captures test results, screenshots, and logs → 
Stores all data in artifact bucket for review

User: "Run the UI tests for the login flow and save the results"

Agent: [Uses Playwright MCP toolkit to execute test cases]
[Creates "ui_test_results_2026-01-09.json" using createFile]
[Creates "test_execution_log.txt" using createFile]
[Saves screenshots for failed tests]

Agent: "I've completed the UI test execution for the login flow:

Test Summary:
- Total Tests: 15
- Passed: 13
- Failed: 2
- Duration: 2m 34s

Failed Tests:
1. Login with invalid credentials - Expected error message not displayed
2. Password reset flow - Email field validation failed

All results have been saved to your artifacts:
- ui_test_results_2026-01-09.json (detailed test results)
- test_execution_log.txt (execution logs)
- screenshot_login_error.png (failed test screenshot)
- screenshot_password_reset.png (failed test screenshot)"

User: "Append a summary to the execution log"

Agent: [Uses appendData to add summary]
"Summary appended to test_execution_log.txt:

=== Test Execution Summary ===
Date: 2026-01-09
Environment: Staging
Browser: Chrome 120
Pass Rate: 86.7%

Action Required:
- Investigate error message display logic in login component
- Review email validation regex in password reset form"

User: "Run the checkout flow tests and save results to the same log"

Agent: [Executes checkout tests using Playwright MCP toolkit]
[Uses appendData to add results to existing log]
[Creates new test results file]

Agent: "Checkout flow tests completed:

Test Summary:
- Total Tests: 8
- Passed: 8
- Failed: 0
- Duration: 1m 47s

Results appended to test_execution_log.txt
Detailed results saved to ui_test_results_checkout_2026-01-09.json

All tests passed! ✓"

​

Troubleshooting and Support

​

Troubleshooting

​

Support Contact

​

Frequently Asked Questions