from agent_memory_bench import run_benchmark, Task, AgentConfig

tasks = [Task("fix_bug_and_preserve_edits", repo="my/example_repo.git")]
agent_configs = [
    AgentConfig(agent="bullet", memory="context_window"),
    AgentConfig(agent="bullet", memory="embedding_recall"),
    AgentConfig(agent="mastra", memory="context_window"),
    AgentConfig(agent="mastra", memory="embedding_recall")
]

results = run_benchmark(tasks, agent_configs, model="gpt-4o", pricing="june_2024")
print(results)