SAGE Agentic Tool Use Benchmark

Configuration-driven experiment framework for evaluating tool selection and use capabilities.

Features

Tool Selection Evaluation: Tool retrieval and ranking benchmarks
Planning Evaluation: Multi-step planning with tool composition
Timing Detection: Timing judgment for tool invocation decisions

Quick Start

# Install
pip install isage-agentic-tooluse-benchmark

# Run tool selection experiment
sage-agentic-tooluse-bench tool-selection --config config/tool_selection_exp.yaml

# Run planning experiment
sage-agentic-tooluse-bench planning --config config/planning_exp.yaml

Documentation

See benchmark_agent/README.md for detailed documentation.

Development

# Clone
git clone https://github.com/intellistream/sage-agentic-tooluse-benchmark.git
cd sage-agentic-tooluse-benchmark

# Setup virtual environment
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install in development mode
pip install -e ".[dev]"

# Run tests
pytest

License

MIT License - see LICENSE for details.

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

isage_agentic_tooluse_benchmark-0.1.7.tar.gz (124.7 kB view details)

Uploaded Mar 4, 2026 Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

The dropdown lists show the available interpreters, ABIs, and platforms. Enable javascript to be able to filter the list of wheel files.

isage_agentic_tooluse_benchmark-0.1.7-py2.py3-none-any.whl (144.4 kB view details)

Uploaded Mar 4, 2026 Python 2Python 3

File details

Details for the file isage_agentic_tooluse_benchmark-0.1.7.tar.gz.

File metadata

Download URL: isage_agentic_tooluse_benchmark-0.1.7.tar.gz
Upload date: Mar 4, 2026
Size: 124.7 kB
Tags: Source
Uploaded using Trusted Publishing? No
Uploaded via: twine/6.2.0 CPython/3.11.14

File hashes

Hashes for isage_agentic_tooluse_benchmark-0.1.7.tar.gz
Algorithm	Hash digest
SHA256	`64d98becc419e196c57a4d889bbcbce1bc396f1d9589a77b510e549d2e13d0dc`
MD5	`b7160f13bc193e45800567b4d4fc73d3`
BLAKE2b-256	`5a562ed38f2eb1e02e80a09606f52661b05f862a11c67efc726838983bb528f1`

See more details on using hashes here.

File details

Details for the file isage_agentic_tooluse_benchmark-0.1.7-py2.py3-none-any.whl.

File metadata

Download URL: isage_agentic_tooluse_benchmark-0.1.7-py2.py3-none-any.whl
Upload date: Mar 4, 2026
Size: 144.4 kB
Tags: Python 2, Python 3
Uploaded using Trusted Publishing? No
Uploaded via: twine/6.2.0 CPython/3.11.14

File hashes

Hashes for isage_agentic_tooluse_benchmark-0.1.7-py2.py3-none-any.whl
Algorithm	Hash digest
SHA256	`f2d6cf6878cf64e1b0316f427dd3141f30becf1bcc716f31a7b4e6f5331e230f`
MD5	`9ed058bb64d8409240fd71c907936fd4`
BLAKE2b-256	`9e2265ed3f8cb6cfa3fddefa2af04c044ad2b17a090986a666adbe19823dcec0`