Crate aptitude 

aptitude

A fluent assertion library for testing AI agent tool calls.

This library provides a Jest-like API for asserting on tool calls made by AI coding agents. It can be used with Rust’s native #[test] framework.

Quick Start

use aptitude::{prompt, expect, Tool};

#[test]
fn test_agent_behavior() {
    let tool_calls = prompt("Read the config file").run().unwrap();

    expect(&tool_calls)
        .tool(Tool::Read)
        .to_be_called();

    expect(&tool_calls)
        .tool(Tool::Bash)
        .not_to_be_called();
}

With Working Directory

use aptitude::{prompt, expect, Tool};

#[test]
fn test_in_project_dir() {
    let tool_calls = prompt("List files here")
        .in_dir("/path/to/project")
        .run()
        .unwrap();

    expect(&tool_calls)
        .tool(Tool::Bash)
        .to_be_called();
}

Analyzing Existing Sessions

use aptitude::{parse_session, expect, Tool};

#[test]
fn test_existing_session() {
    let tool_calls = parse_session("session.jsonl").unwrap();

    expect(&tool_calls)
        .tool(Tool::Write)
        .after(Tool::Read);
}

Re-exports

pub use fluent::expect;

pub use fluent::expect_tools;

pub use fluent::params_match;

pub use fluent::ExecutionExpectation;

pub use fluent::StdoutAssertion;

pub use fluent::ToolAssertion;

pub use parser::parse_jsonl_file as parse_session;

pub use parser::ToolCall;

pub use fluent::Tool;

pub use agents::AgentHarness;

pub use agents::AgentType;

pub use agents::ExecutionConfig;

pub use agents::ExecutionOutput;

pub use agents::NormalizedResult;

pub use prompt::prompt;

pub use prompt::PromptBuilder;

pub use output::OutputConfig;

pub use output::OutputFormatter;

pub use output::OutputMode;

pub use yaml::load_test;

pub use yaml::run_yaml_test;

pub use yaml::Assertion;

pub use yaml::Test as YamlTest;

Modules

agents

Agent abstraction layer for multi-agent support.

config

Configuration file support for aptitude.

discovery

Test file discovery using glob patterns and walkdir.

fluent

Fluent assertion API for testing AI agent execution output.

output

Output formatting for test results, tool calls, and agent responses.

parser

prompt

Prompt builder for fluent prompt configuration and execution.

yaml

YAML test file support for the agent harness.

Macros

params

Create a parameter map from key-value pairs.