Types of Tools
Arcade offers two types of :
The distinction is merely a matter of how they are designed. Both types of can be used seamlessly in the same way. There is no difference in their interfaces, the way they are called, or how you interact with them through the Arcade Dashboard or the Arcade SDK clients.
Before we understand the two types, let’s first understand the background for why we need to differentiate between them.
Why LLMs perform poorly when calling HTTP APIs
Traditionally, the HTTP APIs offered by upstream services such as GitHub, Google, Slack, etc., were designed to be consumed by human software engineers. When we expose such interfaces for LLMs to call as , they usually do not perform very well.
One of the main reasons is that the data model of the HTTP API rarely matches the data model of an AI-powered chat interface.
For instance, consider the following prompt:
“Send a DM to John asking about a update”
The data model mismatches are:
| Dimension | Chat interface | Slack HTTP API |
|---|---|---|
| Action | Send message to a person | Send message to a channel |
| Argument | username = "John" | channel_id = ??? |
In order to bridge the gap in the data models, the LLM has to make multiple API calls:
- Retrieve the current ’s Slack ID
- Browse the list of to find John’s ID
- Open a DM (direct message) channel between the and John, and get this channel’s ID
- Send the message to the channel
Even the most powerful LLMs usually perform poorly when they need to reason such complex workflows on the fly, not to mention the increased cost and risk of hallucinations. As a result, AI and chatbots that rely on HTTP APIs often end up being unreliable.
Optimized tools
Arcade’s Optimized Servers are designed to match the typical data models expected in AI-powered chat interfaces and are subject to evaluation suites to ensure LLMs can safely use them.
Following the example above, our Slack Server offers the Slack.SendMessage , which accepts a username as argument, matching exactly both the action and argument value expected to be present in the LLM window.
When a user says “Send a DM to John asking about a update”, the LLM can directly call the Slack.SendMessage with the username argument, and the tool will take care of the rest.
dramatically improve the speed, reliability and cost-effectiveness of AI and chatbots.
Since they require careful design and evaluation, Optimized tools take time and effort to build. We understand that your Agent or chatbot project might need capabilities not yet covered by our Optimized Servers. For this reason, we also offer low-level Starter .
Starter tools
To provide your Agent or chatbot with more freedom to interact with the upstream services, we offer Starter Servers.
are heavily influenced by the original API design. Each mirrors one HTTP endpoint.
Although we redesign the tool name and argument descriptions to make them more suitable for LLMs, Starter tools are still not optimized for LLM usage. Also, they are not subject to evaluation suites like . For those reasons, we recommend thoroughly evaluating each with your or chatbots before using it in production.
When your Agent’s needs are covered by an , we recommend using it instead of a Starter one. Use as a complement. Carefully engineer your prompts to ensure your can call them safely.