Threads & context

Set trace-level context when you open the trace.

Threads

When one user has a back-and-forth with your agent, each turn is its own trace. Give related turns the same threadId so Lemma groups them into one conversation.

TypeScript
Python

await lemma.trace(
  {
    name: "support-agent",
    input: userMessage,
    threadId: conversationId,
  },
  async (trace) => {
    return runAgent(userMessage, trace);
  },
);

async def run(trace):
    return await run_agent(user_message, trace)

await lemma.async_trace(
    "support-agent",
    run,
    input=user_message,
    thread_id=conversation_id,
)

Users

TypeScript
Python

await lemma.trace(
  {
    name: "support-agent",
    input: userMessage,
    threadId: conversationId,
    userId: user.id,
  },
  async (trace) => {
    return runAgent(userMessage, trace);
  },
);

async def run(trace):
    return await run_agent(user_message, trace)

await lemma.async_trace(
    "support-agent",
    run,
    input=user_message,
    thread_id=conversation_id,
    user_id=user.id,
)

Reference

Context	SDK option	Attribute keys
Agent name	`name`	`gen_ai.agent.name`, `ai.agent.name`
Thread / conversation	`threadId`	`lemma.thread_id`
User	`userId`	`user.id`, `enduser.id`

See the trace contract for how each field is read.

Next steps

Trace contract

The exact shape Lemma reads.

Good trace vs bad trace

Conformant vs malformed traces.

​Threads

​Users

​Reference

​Next steps