Skip to content

Releases: DavidWallstructurallaw/structdet-code

StructDet Code v0.1.0

Choose a tag to compare

@DavidWallstructurallaw DavidWallstructurallaw released this 24 Sep 06:56
95b2be6

StructDet Code 0.1.0

An offline Python CLI for code diversity, algorithmic diversity and recorded coding-agent trajectory analysis.

Included

  • Passive Python source preparation, narrow sorting/graph mechanism rules and source-bound review.
  • Compatible condition comparisons, revision trajectories, checkpoint cohorts and intervention/failure evidence.
  • Exact offline JSON/Markdown replay and six bundled demonstrations.
  • Installable wheel and source archive, including both task packs, all 58 example files and license notices.

Install and try

Use Linux with CPython 3.12.14, the qualified environment. Download the wheel and SHA256SUMS from the assets below, then run:

sha256sum --check --ignore-missing SHA256SUMS
python3.12 -m venv .venv
.venv/bin/python -m pip install --no-index --no-deps ./structdet_code-0.1.0-py3-none-any.whl
.venv/bin/structdet-code demo --output ./first-demo

Choose a new output directory. Open first-demo/summary.md to explore the six replayed analyses. There are no third-party runtime dependencies.

Verification and scope

C06 passed 142 regression methods. The final wheel and actual source archive each passed 16 installed workflow groups (34 business CLI commands per route), outside the checkout. Both produced the same six demonstration results and passed exact offline replay. Runtime, task and example bytes match the original C06 qualification.

Current tasks are bounded integer sorting and minimum-hop distances on directed unit-edge graphs. Unfamiliar code remains unresolved; supplied provenance is not authenticated. The CLI does not execute candidate programs or call models. Other Python versions and operating systems are unqualified. Historical dev5 snapshots require dev5; create fresh snapshots with 0.1.0.

See the repository README, docs/INSTALL.md, docs/RELATED_WORK.md and verification/release_0_1_0.json for details. Python code uses Apache-2.0; documentation and designated data use CC BY 4.0. No PyPI upload is included.