Populate DOCX Contract Templates with Data

This skill automates filling Word document templates - contracts, forms - with structured data using the docxtpl library, which reads Jinja2-style {{VARIABLE_NAME}} placeholders embedded in a .docx file. Its workflow is explicitly script-driven rather than reading the document directly: run scripts/extract_schema.py against the uploaded template to get the list of variables and a JSON schema (the skill is instructed not to try reading the .docx file itself, since docxtpl's schema extraction handles that reliably); gather values for each schema field from the user's message context, using Claude's own extraction ability where needed; explicitly ask the user for any required field that's missing or uncertain rather than guessing or inventing a plausible-sounding value; write the gathered data to a JSON file and run scripts/fill_template.py with the template, the data file, and an output path to produce the filled document; and finally move the result to /mnt/user-data/outputs/ and provide a download link, again without reading the output file directly.

Its key constraints: the template must use Jinja2 {{VARIABLE_NAME}} syntax for docxtpl to recognize placeholders, every required field identified in the extracted schema must be filled before the document is considered complete, and missing data is always a reason to ask the user - never to fill in an invented value. The docxtpl library itself may need installing (pip install docxtpl --break-system-packages) if not already present.

Use this skill when a user uploads a .docx template containing {{variable}} placeholders and provides, or can provide, the data to fill them, or explicitly requests contract or form generation from an existing template.

It is not a general document-editing or contract-drafting tool - it fills an existing template's defined variables rather than writing new contract language or restructuring a document's content, and it depends on the template already using the correct Jinja2 placeholder syntax; a template without {{}} placeholders isn't something this skill's extraction step can work with.

Input is a .docx template with {{VARIABLE_NAME}} placeholders plus the data to fill each required field, extracted from context or supplied directly by the user when missing. Output is a completed .docx document with all template variables populated, delivered as a download link from /mnt/user-data/outputs/.

python scripts/extract_schema.py template.docx

Built on the docxtpl Python library for Jinja2-based Word template filling, using companion scripts extract_schema.py (schema extraction) and fill_template.py (template population).

Users generating contracts or forms from a pre-built Word template who need structured data filled in reliably, with missing fields flagged for user input rather than silently guessed.