Use when a user provides a settlement summary markdown or text file and wants to generate the fiscal appropriation detail Excel workbook from the bundled template, including evidence JSON or regression checks.
Use this skill when the user provides a settlement summary in Markdown or text and asks to create the fiscal appropriation detail spreadsheet, validate the extraction, or run the bundled sample regressions.
Primary input:
.md or .txt format.Primary output:
.xlsx workbook based on assets/template.xlsx.assets/template.xlsx. Never ask the user to upload the blank template again unless they explicitly want to replace the bundled asset.assets/template.xlsx directly.scripts/fill_fiscal_appropriation_detail.py rather than reimplementing the rules manually.Parse to evidence JSON:
python scripts/fill_fiscal_appropriation_detail.py parse path/to/summary.md -o path/to/evidence.json
Fill a workbook:
python scripts/fill_fiscal_appropriation_detail.py fill path/to/summary.md -o path/to/output.xlsx --evidence-json path/to/evidence.json
Fill with a frozen cutoff date for testing:
python scripts/fill_fiscal_appropriation_detail.py fill path/to/summary.md -o path/to/output.xlsx --evidence-json path/to/evidence.json --as-of-date 2026-04-07
Convert bundled fixtures and refresh template assets:
python scripts/convert_fixtures.py
Run bundled regression tests:
python scripts/run_regression.py --as-of-date 2026-04-07
Run unit tests:
python -m pytest tests/test_fiscal_appropriation_detail.py
assets/template.xlsx: bundled blank workbook template used for all production fills.assets/template-fingerprint.json: generated asset provenance and hashes.references/fixtures/: bundled sample summaries and reference answer workbooks.references/field-rules.md: field mapping and extraction rules.references/template-structure.md: workbook layout and row-writing rules.references/acceptable-differences.md: regression difference categories.references/executable-prompt.md: user-facing executable prompt and operating specification.