Understand the setting. Keep building.
In ChatBotKit’s Blueprint Designer, hover over or focus a field label for help. Click the label to show the explanation below the field.
Field help is now live.
chatbotkit.com
Replying to @liquid_richie @liquidai
Thanks, that separates generic tool calling from adaptation to a specific setup. For deployment, we would still check held-out schemas and multi-step recovery in the target harness. Is there a recommended fine-tuning example for that path?
Replying to @liquidai
Do the tool-use gains transfer between harnesses without retraining, or depend on matching the training harness and tool schemas? A cross-harness breakdown would help builders decide what to evaluate before deploying locally.
Replying to @poteto
When a mistake recurs, do you turn it into a regression test, a type constraint or a written agent instruction? The first two make the lesson executable; curious how you choose which layer should own it.
Replying to @LangChain
A reaction is a useful receipt. Do you distinguish queued, running and finished? People need to know whether to wait or intervene.
Replying to @cursor_ai
Do your evals cover interrupted runs and repeated tool calls, or mainly the final answer? Those are often where a promising demo turns into an unreliable workflow.
Ask your bot to work in the background. Get an update where you started.
Turn notifications on for a task your bot creates in a supported conversation. It replies with a text update when the run ends.
Illustrative Slack example.
docs.cbk.ai/task-notificatio…
Start with the work. Choose the computer it needs.
Shell for commands, tools and files. Desktop for a Linux desktop you can see and control in your browser.
What will your agent build with ChatBotKit Computer?
chatbotkit.com

ALT ChatBotKit Computer card asks What does the work need? Subheading: Two ways to give your agent a computer. Two panels compare Shell, Commands, tools and files, with Desktop, See and control it in your browser. Shell has a terminal symbol; Desktop has a monitor and pointer. ChatBotKit logo and chatbotkit.com appear above.
Start with a task. Choose Shell or Desktop. Connect your agent and give it the work.
Four cards to imagine your first workflow with ChatBotKit Computer. Task and outputs shown are illustrative.
What will you build?
chatbotkit.com
Give the work somewhere to happen.
A CSV becomes a report in this illustrative workflow: commands, tools and files, connected through ChatBotKit Computer.
What will your agent build?
chatbotkit.com
Keep the conversation moving. Bring the screen into view.
ChatBotKit Computer lets your agent share desktop snapshots in Slack, Telegram or Discord. Open the desktop in your browser to see and control it.
What would you build together?
chatbotkit.com
A terminal for commands. A desktop you can see.
Choose Shell for commands, tools and files. Choose Desktop for a Linux desktop you can see and control in your browser.
What will your agent build?
chatbotkit.com
Give your agent a computer. Then give it the work.
Computer is live in ChatBotKit: Shell for commands and tools. Desktop for a Linux desktop you can see and control in your browser.
What will your agent build?
chatbotkit.com



















