Quick Start: Connect Toolkits

​

What are Toolkits?

Toolkits connect ELITEA to external services, enabling you to:

• Fetch data from project management tools
• Search documentation
• Analyze code repositories
• Read design files
• Manage test cases
• Store and retrieve files

---

​

Before You Begin

---

​

Jira Toolkit

Connect to Jira for issue tracking, project management, and agile workflows. How to Connect

1. Access Chat: Click Chat in the left sidebar
2. Create a Toolkit from Chat:
   • In the PARTICIPANTS panel on the right, find toolkits
   • Click the + Add toolkit icon
   • Click + Create new Toolkit
   • Select Jira as the toolkit type
   • Enter a name (e.g., MyJiraToolkit)
3. Create Credentials:
   • In the CONFIGURATION section, click the credentials dropdown
   • Select + New credentials (or + New private credentials)
   • You’ll be redirected to the Create Credentials page
   • Select Jira as the credential type
   • Give it a name like My Jira Credentials
   • Fill in:
     • Base URL: Your Jira instance URL
     • For Basic Auth: Username + API Key
     • For Bearer Token: Token only
   • Click Save
4. Complete Toolkit Configuration:
   • Return to the toolkit canvas (or refresh if needed)
   • Click the refresh icon next to the Jira configuration field
   • Select your newly created credential from the dropdown
   • Select an Embedding Model (e.g., text-embedding-3-small)
   • Choose a Vector Store (PostgreSQL with pgvector extension)
   • Enable the tools you need (e.g., Search using JQL, Create issue)
   • Click Create
   • Close the Canvas interface
5. Use it:
   • Type: Find all open bugs in project ABC
   • The toolkit will search Jira and return results

---

​

Confluence Toolkit

Connect to Confluence for documentation search and knowledge base access. How to Connect

1. Access Chat: Click Chat in the left sidebar
2. Create a Toolkit from Chat:
   • In PARTICIPANTS panel, click + Add toolkit
   • Click + Create new Toolkit
   • Select Confluence
3. Create Credentials:
   • Click credentials dropdown → + New credentials
   • Select Confluence as credential type
   • Name it (e.g., My Confluence Credentials)
   • Fill in:
     • Base URL: Your Confluence instance URL
     • For Basic Auth: Username + API Key
     • For Bearer Token: Token only
   • Click Save
4. Complete Toolkit Configuration:
   • Return to toolkit canvas
   • Refresh and select your credential
   • Specify a Confluence space to limit searches-the toolkit will be automatically named with this space name
   • Select an Embedding Model (e.g., text-embedding-3-small)
   • Choose a Vector Store (PostgreSQL with pgvector extension)
   • Enable the tools you need (e.g., Read page by id, Update page by title)
   • Click Create
5. Use it:
   • Type: Search Confluence for API documentation
   • The toolkit will search and return relevant pages

---

​

GitHub Toolkit

Connect to GitHub repositories for code analysis, issue management, and pull request operations. How to Connect

1. Access Chat: Click Chat in the left sidebar
2. Create a Toolkit from Chat:
   • In PARTICIPANTS panel, click + Add toolkit
   • Click + Create new Toolkit
   • Select GitHub
3. Create Credentials (optional for public repos):
   • Click credentials dropdown → + New credentials
   • Select GitHub as credential type
   • Name it (e.g., My GitHub Credentials)
   • Fill in (choose one method):
     • Token: Access Token only
     • Password: Username + Password
     • App Private Key: App ID + Private Key
   • Click Save
4. Complete Toolkit Configuration:
   • Return to toolkit canvas
   • Refresh and select your credential (or leave empty for public access)
   • Configure:
     • Project: Organization or username
     • Repository: Repository name(e.g., ProjectAlita/projectalita.github.io)
     • Branch: Default branch (e.g., main)
   • Select an Embedding Model (e.g., text-embedding-3-small)
   • Choose a Vector Store (PostgreSQL with pgvector extension)
   • Enable the tools you need (e.g., Read file, List pull request diffs)
   • Click Create
5. Use it:
   • Type: Get the guide from projectalita.github.io repo for creating secrets in ELITEA
   • The toolkit will fetch and display the secret creation documentation
   

---

​

Figma Toolkit

Connect to Figma for design file access and collaboration. How to Connect

1. Access Chat: Click Chat in the left sidebar
2. Create a Toolkit from Chat:
   • In PARTICIPANTS panel, click + Add toolkit
   • Click + Create new Toolkit
   • Select Figma
   • Enter a name (e.g., MyFigmaToolkit)
3. Create Credentials:
   • Click credentials dropdown → + New credentials
   • Select Figma as credential type
   • Name it (e.g., My Figma Credentials)
   • Fill in:
     • Token: Your Figma Personal Access Token
   • Click Save
4. Complete Toolkit Configuration:
   • Return to toolkit canvas
   • Refresh and select your credential
   • Select an Embedding Model (e.g., text-embedding-3-small)
   • Choose a Vector Store (PostgreSQL with pgvector extension)
   • Enable tools for accessing Figma files(e.g. Get file, Get project files)
   • Click Create
5. Use it:
   • Type: Get information from Figma file XYZ
   • The toolkit will access Figma and retrieve file data
   

---

​

TestRail Toolkit

Connect to TestRail for test case management and test execution tracking. How to Connect

1. Access Chat: Click Chat in the left sidebar
2. Create a Toolkit from Chat:
   • In PARTICIPANTS panel, click + Add toolkit
   • Click + Create new Toolkit
   • Select TestRail
3. Create Credentials:
   • Click credentials dropdown → + New credentials
   • Select TestRail as credential type
   • Name it (e.g., My TestRail Credentials)
   • Fill in:
     • URL: Your TestRail instance URL (e.g., https://yourcompany.testrail.io)
     • Email: Your account email (e.g., user@company.com)
     • Password: Your password or API key
   • Click Save
4. Complete Toolkit Configuration:
   • Return to toolkit canvas
   • Refresh and select your credential
   • Enter a name (e.g., MyTestRailToolkit)
   • Select an Embedding Model (e.g., text-embedding-3-small)
   • Choose a Vector Store (PostgreSQL with pgvector extension)
   • Enable the tools you need (e.g., Get cases, Add case)
   • Click Create
5. Use it:
   • Type: Show me all test cases related to log-in functionality from project with id 1
   • The toolkit will fetch and display test case information
   

---

​

Artifact Toolkit

Create storage buckets for file attachments in chat conversations. How to Connect Manual Setup from Toolkit Canvas:

1. Access Chat: Click Chat in the left sidebar
2. Create a Toolkit:
   • In PARTICIPANTS panel, click + Add toolkit
   • Click + Create new Toolkit
   • Select Artifact
   • Enter a name that matches your bucket name
3. Configure Bucket:
   • Bucket: Enter a unique name (lowercase, alphanumeric, hyphens)
     • ✔️ Valid: my-files, project-2024, team-assets
     • ✘ Invalid: My Files, project_2024, Team@Assets
   • Select an Embedding Model (e.g., text-embedding-3-small)
   • Choose a Vector Store (PostgreSQL with pgvector extension)
   • Enable the tools you need (e.g., CreateFile, ListFiles)
   • Click Create
4. Use it:
   • Files uploaded via chat are stored in this bucket
   • Agents can read, create, and manage files programmatically
   • Example: List all files in my artifact bucket or Read the contents of report.pdf
   

---

​

Troubleshooting

Authentication Errors (401/403)

• Verify credentials are correct
• Check API token hasn’t expired
• Ensure user has required permissions
• For Atlassian tools, verify correct authentication method

Connection Issues

• Verify base URL is correct** (must include http:// or https://)
• Check firewall/proxy settings
• Test connection from browser first
• Ensure service is accessible from your network

Toolkit Not Appearing

• Click refresh icon next to credential selector
• Verify toolkit was created successfully
• Check project permissions
• Reload the page

Tools Return No Results

• Check the “Limit” field in toolkit configuration
• Verify search parameters are correct
• Ensure you have access to the requested data
• Check service-specific query syntax (e.g., JQL for Jira)

---

---