dataforge-07-agent-patterns 0.1.0

Reusable agentic primitives extracted from DataForge.

dataforge-agent-patterns

Reusable pure-Python primitives for agent projects:

from dataforge_agent_patterns import ProgressiveToolDisclosure

disclosure = ProgressiveToolDisclosure({"search": 1, "shell": 3})
assert disclosure.visible_tools(task_difficulty=2) == ("search",)

The package is intentionally independent from dataforge. Runtime imports are limited to the Python standard library, pydantic, and lazily imported z3 inside SMTVerifiedAction. It is not published to PyPI yet; use the local verification source install until Trusted Publishing ownership is configured.

Primitives

• ProgressiveToolDisclosure - reveal tools only when task difficulty warrants them.
• ConstitutionalFilter - wrap an agent with a safety-verdict layer.
• ReversibleTransaction - decorate side-effect calls with rollback support.
• SMTVerifiedAction - check structured actions with Z3 before execution.
• CausalCascadeDetector - flag cascading effects in action graphs.

Local Verification

python -m pip install -e ".[dev]"
python -m ruff check src tests
python -m ruff format --check src tests
python -m mypy src
python -m pytest
python -m build

PyPI publishing should use GitHub Actions Trusted Publishing after a pending publisher is configured for dataforge_07_agent_patterns; do not add PyPI API tokens.