Quickstart

Install

pip install staso

To upgrade to the latest version:

pip install -U staso

Python 3.11+ required.

Initialize

Get your API key from the Staso dashboard, then:

import staso as st

st.init(
    api_key="ak_...",
    agent_id="my-agent",
    environment="production",
)

Add Decorators

Put @st.agent on your agent's entry-point. Put @st.tool on its tools. Done.

@st.agent(name="support-agent")
def handle_request(user_message: str) -> str:
    context = search_faq(user_message)
    return generate_response(context, user_message)

@st.tool(name="search_faq")
def search_faq(query: str) -> str:
    return db.search(query)

Every call to handle_request now shows up on your dashboard — which tools ran, what they returned, how long each step took, and any errors.

Group by Conversation

Tie multiple agent runs into a conversation:

with st.conversation("conversation-123"):
    handle_request("How do I reset my password?")
    handle_request("That didn't work")

Shut Down

Flush remaining traces before your process exits:

st.shutdown()

Full Example

import staso as st

st.init(api_key="ak_...", agent_id="support-agent", environment="production")
st.integrations.patch_anthropic()  # auto-trace Anthropic calls

@st.tool(name="search_faq")
def search_faq(query: str) -> str:
    return "To reset your password, go to Settings > Security > Reset Password."

@st.agent(name="support-agent")
def run_agent(user_message: str) -> str:
    import anthropic
    faq = search_faq(user_message)
    client = anthropic.Anthropic()
    response = client.messages.create(
        model="claude-sonnet-4-20250514",
        max_tokens=1024,
        messages=[{"role": "user", "content": f"Context: {faq}\n\nQuestion: {user_message}"}],
    )
    return response.content[0].text

with st.conversation("customer-jane-001"):
    print(run_agent("How do I reset my password?"))

st.shutdown()

On the dashboard you'll see:

support-agent (agent) — total duration, status
search_faq (tool) — input query, returned result
anthropic.messages.create (llm) — model, token usage, latency

Tracing your code — @agent, @tool, @trace
Conversations — Conversation grouping
LLM integrations — Auto-instrument Anthropic, OpenAI
Claude Code — Trace Claude Code CLI conversations
Configuration — All options