Small decisions. Interesting possibilities.Submit contentSubmit

A Slack agent, twice as fast

Jev made our Slack agent 2x faster ⚡️ Our agent can be quite slow because it needs to read skills and figure out which tools to call. We used @typesafeai's new model to speed this up by first passing it the prompt and classifying the best skill, tool and params to use before handing it to the agent

Added to Jevfast

How it uses Jev

Jev classifies the skill, tool, and parameters for a prompt before handing it to the Slack agent.

App workflow

  1. Receive the prompt

    The user prompt is passed to Jev before the agent handles it.

  2. Classify the route

    Jev identifies the best skill and tool for the prompt.

  3. Supply parameters

    The classification includes the parameters to use.

  4. Hand off to the agent

    The selected skill, tool, and parameters are passed on to the agent.

Why it is interesting

The router tries to select a skill, tool, and parameters before the Slack agent reads a larger tool set. It is a concrete attempt to shorten an agent’s planning step with a typed decision.