Agent Assertions

Validate tool/function calling behavior.

tool_called


expect.tool_called("web_search")           # called at least once
expect.tool_called("web_search", times=1)  # called exactly once
expect.tool_called("web_search", min_times=2)  # at least 2 times

no_loop_detected

Detects infinite loops in tool call traces (same tool called 3+ times consecutively).


expect.no_loop_detected()

Loop detection looks for the same tool called 3+ times in a row. This catches runaway agents before they burn through your budget.

tool_call_order

Checks tools were called in order.


expect.tool_call_order(["search", "parse", "summarize"])

Full Example

test_agent.py


from assertllm import expect, llm_test
 
@llm_test(
    expect.tool_called("search", min_times=1),
    expect.tool_call_order(["search", "summarize"]),
    expect.no_loop_detected(),
    expect.latency_under(5000),
    model="claude-sonnet-4-6",
)
def test_research_agent(llm):
    tools = [
        {
            "type": "function",
            "function": {
                "name": "search",
                "description": "Search the web",
                "parameters": {
                    "type": "object",
                    "properties": {"query": {"type": "string"}},
                },
            },
        },
        {
            "type": "function",
            "function": {
                "name": "summarize",
                "description": "Summarize text",
                "parameters": {
                    "type": "object",
                    "properties": {"text": {"type": "string"}},
                },
            },
        },
    ]
    llm("Find and summarize recent AI news", tools=tools)

Output


test_agent.py::test_research_agent
  Tool calls: search → summarize
  ✓ tool_called("search", min_times=1) — called 1 time
  ✓ tool_call_order(["search", "summarize"])
  ✓ no_loop_detected()
  ✓ latency_under(5000) — 2341ms
  PASSED    [2.3s]