Experience-Driven Black-Box Cost Control for Coding Agents
Python agentic-aiagentic-codingai-codingai-coding-assistantanthropicartificial-intelligenceclaude-codecodexcoding-agentcoding-agentscost-optimizationcost-reductiondeveloper-toolsgithub-copilotllmopen-sourceopenaiproductivitytoken-optimization
# TokenEconomizer — repo-layer requirements (Layer 1).
#
# The CORE package (`import tokeneconomizer`) is STDLIB-ONLY and needs nothing here.
# This pin covers the one optional repo capability: running the test suite. It mirrors the
# [test] extra in pyproject.toml.
#
# pip install -r requirements.txt # tests
# pip install -e . # then install the package itself (editable)
#
# NOT included on purpose (these belong to the EXTERNAL harness, Layer 2 — see
# README.md): Docker, the sealed agent CLIs, and the
# provisioning-only tools (git, Node/npm, the HuggingFace `datasets` package).
pytest>=7.0 # offline test suite (no Docker, no datasets)