Add scientific research workflows plugin

[hetaoBackend]

What changes

Adds hetaoBackend/scientific-research-workflows, a MiniMax Code Plugin that curates nine
MIT-licensed Skills from K-Dense's Scientific Agent Skills at pinned commit
1dd0fccf46fc3c9855c4a0c313a0c57fe4319883:

• scientific brainstorming and hypothesis generation;
• experimental design, statistical power, statistical analysis, and uncertainty/units;
• scientific writing, peer review, and scholarly evaluation.

The package includes each selected Skill's references, assets, and local scripts. It excludes
network-dependent and separately licensed upstream Skills. The scientific-writing draft placeholder
label is renamed consistently to satisfy this repository's hosted-package validator; the README
records that adaptation and the exact upstream snapshot.

User value

Researchers can use one coherent Plugin to move from an early idea to a defensible study design,
analysis plan, manuscript, and review while keeping bundled tooling local-first.

Example prompt:

I am planning a two-site experiment comparing three treatments, but samples arrive in four weekly
batches. Help me define competing hypotheses, design randomization and blocking, estimate the
sample size for 80% power, and produce a preregistration-ready analysis outline. State every
assumption and do not invent pilot data.

Expected result: the agent separates hypotheses from evidence, identifies experimental units and
confounders, proposes a reproducible blocked design, requests effect-size inputs, and produces an
analysis plan with explicit uncertainty and reporting boundaries.

Plugin submission checklist

• Plugin lives at plugins/<github-owner>/<plugin-name>.
• plugin.json name matches the Plugin directory.
• README.md includes a real example prompt and expected result.
• LICENSE and plugin.json declare an open-source license.
• Required executables, accounts, paid services, and supported platforms are disclosed.
• Network destinations and data handled by the plugin are disclosed.
• No credentials, private endpoints, hidden telemetry, installers, symlinks, or native binaries are included.
• Every scaffold placeholder has been replaced.
• npm run check passes.

Evidence

• npm run check: 27 tests passed; validator accepted 8 hosted Plugins and all examples.
• All 49 bundled Python scripts compiled successfully with Python 3.14.6.
• All 32 dependency-free local CLI help entries completed successfully.
• In an isolated Python 3.13 environment, smoke tests passed for randomization, full and fractional
  factorial design generation, closed-form power, simulation-based power, statistical assumption
  checks, and unit conversion with propagated uncertainty.
• Eight Skill directories match the pinned upstream snapshot byte-for-byte. The scientific-writing
  diff is limited to the documented draft-marker compatibility adaptation.
• No live MiniMax Code UI installation was performed; repository validation and executable CLI
  entry points are covered above.