UCAgentBySoldierBoy

NutShell Cache verification entry for the UCAgent competition.

Python Environment

Use the dedicated Conda environment before running project Python commands:

conda activate ucagent-cache

For non-interactive commands, prefer:

conda run -n ucagent-cache python -m pytest -q

Create or update the environment from environment.yml:

conda env update -n ucagent-cache -f environment.yml

Current Status

P0 scaffold is in progress. The repository currently includes a transaction-level reference model and scoreboard. rtl/Cache.v has been imported from Example-NutShellCache; the next external dependency is the Picker/Verilator toolchain needed by make gen_dut.

Run the P0 Python tests:

conda run -n ucagent-cache python -m pytest -q tests/test_api_basic.py

Check the placeholder interface map structure:

conda run -n ucagent-cache python scripts/check_interface_map.py

Generate current scaffold coverage/report artifacts:

conda run -n ucagent-cache python scripts/write_p0_coverage.py
conda run -n ucagent-cache python scripts/build_report.py

Check whether Picker is available:

picker --version

Current planning documents:

• 网页研读笔记
• 参赛实现方案
• AI 协同记录
• LLM 执行任务书

UCAgent assets:

• NutShell Cache skill
• Prompt set