Skip to main content
ReasonBlocks captures the model calls in your agent’s workflow and uses your application’s tasks to train and evaluate a smaller model. You keep your provider credentials and your existing Python client library throughout. OpenAI, Anthropic, Google Gemini and Amazon Bedrock applications connect through a dashboard data source. Anthropic and OpenAI clients use the stable source URL directly; native Bedrock uses the Python SDK with your existing IAM credentials. Existing generated helpers remain supported. Provider-specific examples are in Integrating your agent.

​
Start with capture

Create a source in Data, generate a persistent server key, and copy the source-preselected setup snippet. Store the key as REASONBLOCKS_API_KEY in your server’s secret environment. Keep the same URL and ReasonBlocks key when moving from recording to an approved, supported release. Your provider credential stays in place. The optional Connection helper and native Bedrock integration use rbtrace==1.3.1; a directly configured Anthropic client needs no helper upgrade. Existing 1.3.0 integrations remain compatible. Version 1.3.1 adds optional Bedrock dispatch diagnostics without changing the connection. Your coding agent can follow the setup guide, also available as /agent-setup.md. Existing generated helpers and their migration workflow remain documented there. Manage connection automatically reports recording, training eligibility, exclusions, failures and recent activity. Capturing does not require a sandbox or start training. Normal provider charges continue when your application runs.

​
Train when your application is ready

Full-agent training needs representative tasks and a test copy of your tools and data that can reset between attempts. Your coding agent can help build the adapter for that environment. ReasonBlocks then prepares a plan, trains within the approved budget and reports complete-task results for review. Enabling a trained release is a separate decision. See Train your complete agent for the environment, evaluation and serving requirements.

​
Get started

Set up with your coding agent

One instruction, a setup CLI and a reusable skill.

Quickstart

Create a data source and connect your existing application.

Integrating your agent

Client settings, task headers and deployment checks.

Train your complete agent

Prepare a test environment, review training and evaluate a release.

Deploy your connected application

Ship the helper, connection configuration and runtime secrets.