docassemble
temaki.docassemble
A Docker Compose file for docassemble alone, generated from the catalog and checked by the safety inspector. No account needed.
Expert system for guided interviews that gathers answers question by question and assembles them into PDF, DOCX or RTF documents, with signatures, conditional logic and multilingual support.
docassemble runs interviews defined in YAML with Markdown question text and Python expressions, using a need-based execution model: rather than following a fixed script, it works out which variables an outcome requires and asks only the questions needed to obtain them. Answers populate Word or PDF templates to produce finished documents, and interviews can equally send email, call APIs, write to a database or file electronically instead of generating a document. Touchscreen signature capture, OCR, background tasks and a browser-based development playground are built in. It is widely used for court forms and legal self-help tools.
You know it worked when
- The web interface loads and the administrator account can sign in.
- The bundled demonstration interview runs to completion.
- An interview authored in the playground collects answers and assembles a PDF.
- A captured signature appears in the generated document.
- Installed packages and sessions persist across a container restart.
Known sharp edges
- The container bundles PostgreSQL, Redis, a background worker and a document converter, so it is a small system in one image and its startup takes minutes rather than seconds.
- The persistent data directory holds uploaded files, the database and installed packages, and losing it destroys interviews and their sessions together.
- Interviews routinely collect sensitive personal information, which means session storage, backups and log retention are a privacy decision rather than an operational detail.
- Document conversion depends on a headless office suite inside the container, and a template using an unsupported construct fails at assembly time rather than at design time.