From LabVIEW to Python

Replace LabVIEW measurement applications with auditable Python software.

An evidence-gated Agent Skill and reusable rule for migrating laboratory and industrial LabVIEW applications to Python without losing measurement behavior, data lineage, hardware safety, operator usability, or packaged-runtime proof.

Independent community project. This repository does not ship vendor SDKs, certify connected instruments, or turn simulator success into a real-hardware claim. Hardware remains unvalidated until an approved fail-closed dry run.

Why this skill

Measurement truth	Operator-ready UI	Delivery evidence
Inventories VI/subVI behavior, resolves one immutable scan plan, and compares workflow, saved, plot, and heatmap data.	Covers industrial and LabVIEW-style interfaces, real clicks and hover, compact/wide layouts, screenshots, and human visual review.	Separates simulator, source UI, packaged EXE, shortcuts, and real hardware; stale binaries and inherited PASS states are rejected.

Real migration results

Pump-probe acquisition	Spectroscopy and TCSPC mapping

These sanitized completion-state screenshots come from real migration verification artifacts. The visual case study explains the demonstrated workflow and the evidence boundary. Textual input/output cases cover supervision, heatmap parity, and packaged-EXE acceptance.

Install

With a compatible Agent Skills installer:

npx skills add k-telux/labview-to-python \
  --skill labview-to-python-industrial

For Codex explicitly:

npx skills add k-telux/labview-to-python \
  --skill labview-to-python-industrial \
  --agent codex

Manual Codex installation:

Copy-Item -Recurse -Force `
  .\skills\labview-to-python-industrial `
  $env:USERPROFILE\.codex\skills\

Quick start

Migrate an application

Use $labview-to-python-industrial to migrate this LabVIEW measurement VI to
Python. Preserve the real scan workflow, add a deterministic simulator, prove
saved and displayed data parity, and package only after source UI approval.

Audit an existing migration

Audit this LabVIEW-to-Python project. Trace every visible parameter into the
resolved scan plan, compare workflow/disk/plot arrays independently, and report
simulator, source UI, EXE, and real hardware as separate statuses.

Diagnose performance or heatmap defects

Use $labview-to-python-industrial to diagnose this slow scan and misregistered
heatmap. Preserve the full workload, measure core and click-to-Completed time,
and verify first/middle/last heatmap cells with real mouse hover.

How it works

Inventory VI/subVI states, controls, units, errors, files, and hardware calls.
Resolve visible controls into one scan plan shared by execution and metadata.
Keep real adapters and deterministic simulators behind the same workflow.
Prove workflow result, saved readback, plots, and heatmaps independently.
Exercise source UI with real interaction and fresh target-window screenshots.
Rebuild and run the exact EXE only after the source gate is approved.

Acceptance gates

Gate	Required proof	Does not prove
Logic and data	Resolved plan, independent fixtures, arrays, masks, units, saved readback	UI, EXE, hardware
Simulator	Deterministic workflow and adapter-boundary behavior	Vendor driver or instrument
Source UI	Real interactions, terminal state, plots, screenshots, visual audit	Packaged runtime
Packaged EXE	Current-source lineage, hashes, shortcuts, real workflows, saved outputs	Real hardware
Real hardware	Approved dry run, limits, communication, cleanup, acquired outputs	Unconnected or optional devices

Quick previews are interaction checks, not production-resolution parity. A target-derived CSV roundtrip is labeled as such and never presented as an independent LabVIEW algorithm oracle.

Example workflows

Case	What it demonstrates
Dual-project supervision	Hard ownership boundaries, shared workflow requirements, and independent phase statuses.
Performance and heatmap gate	Full-workload timing, verifier-induced slowdown, array orientation, and real-hover registration.
Packaged EXE gate	Source freshness, root/dist identity, shortcut integrity, packaged interaction, and output readback.

Practical advantages

No UI-only rewrites: visible controls must affect the executed scan plan.
No self-proof: target-vs-target comparisons and setter-driven hover checks fail.
No status inflation: source, simulator, EXE, and hardware verdicts stay separate.
No silent partial data: stopped scans retain masks and NaN semantics.
No stale delivery: relevant source or verifier changes invalidate old EXEs.
No unsafe process cleanup: ownership is checked before terminating a process.

Repository map

skills/labview-to-python-industrial/   canonical English skill and references
rules/                                 derived project rule summaries
examples/showcase/                     sanitized visual migration case
examples/01..03/                       real-history input/output cases
README.zh-CN.md                        Simplified Chinese entry point
README.ja.md                           Japanese entry point
scripts/validate_repo.py               dependency-free release gate

The English SKILL.md is the technical source of truth. Localized README and rule files are human-facing entry points and defer to the canonical skill when wording differs.

Validate

python scripts/validate_repo.py

Codex users can also run the bundled skill validator:

py $env:USERPROFILE\.codex\skills\.system\skill-creator\scripts\quick_validate.py `
  .\skills\labview-to-python-industrial

Evidence boundary

SDK files and hashes prove inventory, not communication or calibration.
Screenshots prove visible states only when tied to real interaction and data.
Build success is not packaged-runtime success; the exact EXE must be run.
Real hardware requires instrument-specific manuals, safety limits, operator approval, and an actual controlled test.
Published screenshots are sanitized examples, not universal hardware support claims or downloadable applications.

See CONTRIBUTING.md for new adapters, rules, and case studies, and SECURITY.md for responsible disclosure.

Maintained by telux. Released under the MIT License.

Labview2Python

README