AI agent stack
An AI agent stack for building, testing, and shipping agents with tools, memory, browser actions, and evals without turning the setup into a huge mess.
Coding
Loading
This is the stack for teams who are past demos and now need to see what their agents are doing in production, test behavior before launch, and compare model changes without flying blind.
Workflow stack
The order matters. Start at the top, read down the sequence, and open any step when you want the note behind it.
Trace and monitor
Good fit when the team wants open-source-friendly tracing and prompt visibility in one place.
Open tool profileWatch agent sessions
Adds the agent-specific debugging view that gets more valuable once workflows span tools and long-running steps.
Open tool profileGate releases
Useful for benchmark, red-team, and regression testing so the team is not shipping prompt changes by gut feel.
Open tool profileEvaluate behavior
Rounds out the stack with deeper open-source eval and debugging workflows for LLM application quality.
Open tool profileTools in this stack
Open any tool profile if you want pricing, fit, or comparison details.
Unified API for accessing and routing across many language models and providers with one integration layer.
Related stacks
An open-source LLM observability and prompt management platform with free, pro, team, and enterprise tiers.
Observability for AI agents with tracing, debugging, session visibility, and production monitoring.
Evaluation and testing toolkit for prompts, models, and LLM application behavior.
Open-source AI observability tool for tracing, evaluations, experiments, and debugging LLM applications.
Compare tools in this stack
For teams that want internal AI chat and document workflows without handing everything to a public SaaS assistant. This stack leans into self-hosting, local models, and controlled access to stronger remote models when needed.
Official StackBased Editorial Postings
