Skip to content
cccskills
FreePublished on GitHub

solution-package-builder

Assemble a submission-ready writer package from final method, result, robustness, figure, and human-decision artifacts, then freeze approved numerical claims with provenance.

The English summary is being prepared. Showing the original description.

See how to install

Files included (1)

  • SKILL.md2.9 KB

SKILL.md (original)

Review the instructions the agent will receive before you install.

Purpose

Create the writer's single source package and immutable numerical snapshot. Do not manufacture missing judgments or freeze unapproved claims.

Preconditions

  • rigor_profile is submission.
  • Final method explanation, final result analysis, robustness report, and figure plan exist.
  • Human method, result, stability, and claim-scope decisions exist in qx_decisions.jsonl.
  • Canonical sources are current.

Workflow

  1. Collect final method structure, result claims, limitations, figures, tables, and decision provenance.
  2. Draft results/Qx/reports/qx_solution_package_for_writer.md.
  3. Flag every proposed top-line claim with:
    • value and unit;
    • canonical source path and location;
    • robustness support;
    • decision ID for claim scope or rationale;
    • confidence/limitation.
  4. Invoke one final choice card when package sign-off is missing:
    • keep;
    • downgrade;
    • drop.
  5. Route the human answer to modeler-decision-logger as package_signoff.
  6. Only after sign-off, generate results/Qx/reports/frozen_numbers.json.
  7. Verify every numerical package claim resolves to the freeze and every judgment claim resolves to a human decision ID.

Frozen Number Contract

Each entry contains:

{
  "claim_id": "q1_main_rmse",
  "value": 2.4,
  "unit": "units",
  "source_file": "results/Q1/experiments/final/metrics/main.json",
  "source_locator": "$.rmse",
  "frozen_at": "ISO-8601",
  "frozen_by_skill": "solution-package-builder",
  "decision_id": "q1_package_signoff"
}

Use a source line only for stable text files; use a JSON path, table key, or row/column identifier for structured data.

Rules

  • Do not create solution-package-builder_modeler_decision.md.
  • Do not emit frozen_numbers.json before human package sign-off.
  • Never edit an existing freeze by hand.
  • Transcribe human rationales; do not re-compose them as stronger claims.
  • The package may cite the compact method-card history but must not depend on a separate iteration log.
  • Do not copy raw exploratory outputs into the package without final validation.

Change and Re-freeze

When a canonical source changes after freeze:

  1. append the reason to freeze_change_log.md;
  2. update and rerun the canonical source;
  3. obtain renewed human judgment when the evidence changed materially;
  4. regenerate the freeze;
  5. run scoped consistency for Qx.

Verification

  • All prerequisites are final and current.
  • Package sign-off exists in the JSONL ledger.
  • Every frozen value has a real canonical source and stable locator.
  • Every package number matches the freeze.
  • Every modeling judgment traces to a human decision.
  • Writer can use the package without searching scattered results.

Reviews

No reviews yet. Tell others what you thought.

Comments

No login needed. Share how it worked for you, ask a question, or note where it did not run.

No comments yet. Be the first to write one.

Comments are published right away. Do not include personal information such as email addresses or phone numbers. Up to 2 URLs.

0 / 1,000 characters

This form is protected by reCAPTCHA, and the Google Privacy Policy and Terms of Service apply.

More skills from this repository

Detect whether approved modeling code is Python or MATLAB/Beita Tianyuan and route it to the matching reviewer using the compact named-check review contract.

The English summary is being prepared. Showing the original description.

zhnnky329/MathModeling-skills1.1KUpdated Sep 24, 2026

Audit whether the semantic evidence required by the active lean or submission profile exists and is current, without requiring one verbose artifact per skill or an arbitrary number of pass bullets.

The English summary is being prepared. Showing the original description.

zhnnky329/MathModeling-skills1.1KUpdated Sep 24, 2026

Run scoped or final cross-media consistency checks for mathematical-modeling artifacts, comparing canonical numbers, symbols, parameters, decisions, files, and paper claims without performing full-workspace audits for low-risk changes.

The English summary is being prepared. Showing the original description.

zhnnky329/MathModeling-skills1.1KUpdated Sep 24, 2026

Map contest attachments to subquestions, audit and clean raw data, and emit one reusable data profile with quality, coverage, imbalance, concentration, and method-readiness evidence for downstream risk screening.

The English summary is being prepared. Showing the original description.

zhnnky329/MathModeling-skills1.1KUpdated Sep 24, 2026

Build one compact choice card at a genuine mathematical-modeling judgment point. Use before method screening, after a meaningful experiment, or before final claim/freeze approval so the human chooses the trade-off while AI handles mechanical consequences.

The English summary is being prepared. Showing the original description.

zhnnky329/MathModeling-skills1.1KUpdated Sep 24, 2026

Plan the smallest set of diagnostic, comparison, paper, and appendix figures or tables needed to support verified mathematical-modeling decisions and claims.

The English summary is being prepared. Showing the original description.

zhnnky329/MathModeling-skills1.1KUpdated Sep 24, 2026

See all skills by zhnnky329

Report a problem with this skill