Skip to main content
Solace Agent Mesh provides a powerful suite of built-in tools that enable agents to perform various tasks including artifact management, data analysis, multimedia processing, and web research.

Tool Categories

The built-in tools are organized into the following categories:

Artifact Management

Tools for creating, reading, updating, and managing artifacts within agent sessions.

Data Analysis

Tools for analyzing data, creating visualizations, and generating charts.

Audio Tools

Tools for text-to-speech generation, audio transcription, and audio manipulation.

Image Tools

Tools for image generation, editing, and analysis using AI models.

Research Tools

Tools for web search and comprehensive research capabilities.

General Utility Tools

Tools for file conversion, diagram generation, and web requests.

Embeds

A powerful system for dynamic content interpolation and artifact references.

Tool Configuration

Built-in tools are configured in your agent’s YAML configuration file under the builtin_tools section:
builtin_tools:
  - name: list_artifacts
  - name: load_artifact
  - name: text_to_speech
    config:
      gemini_api_key: ${GEMINI_API_KEY}
      model: gemini-2.5-flash-preview-tts
  - name: create_image_from_description
    config:
      model: dall-e-3
      api_key: ${OPENAI_API_KEY}
      api_base: https://api.openai.com

Tool Scopes

Each tool requires specific scopes to be granted to the agent. Scopes control which tools can be accessed:
scopes:
  - tool:artifact:*  # All artifact tools
  - tool:audio:tts   # Text-to-speech
  - tool:image:*     # All image tools
  - tool:web_search:execute

Common Patterns

Using Tools in Sequence

Tools can be chained together to accomplish complex tasks:
# 1. Load data artifact
result = await load_artifact(
    filename="sales_data.csv",
    version=0
)

# 2. Analyze with extract_content_from_artifact
analysis = await extract_content_from_artifact(
    filename="sales_data.csv",
    extraction_goal="Summarize top 5 products by revenue",
    version=0
)

# 3. Create visualization
chart = await create_chart_from_plotly_config(
    config_content=plotly_json,
    config_format="json",
    output_filename="sales_chart.png"
)

Error Handling

All tools return a ToolResult object with status information:
result = await some_tool(...)

if result.status == "error":
    # Handle error
    error_message = result.message
else:
    # Process successful result
    data = result.data

Next Steps

Artifact Management

Learn how to manage files and data within agent sessions

Data Analysis

Explore data analysis and visualization tools

Audio Tools

Generate and process audio content

Image Tools

Create and manipulate images with AI

General Tools

File conversion, diagrams, and web requests

Research Tools

Web search and deep research capabilities

Build docs developers (and LLMs) love