Practical overview
When it fits
Use TrueCourse when a team wants documented product requirements translated into repeatable flows and tests, including checks that can surface mismatches during pull-request review.
Who it is for
- Product and engineering teams that maintain requirements or documentation alongside a software product.
- Teams seeking recurring checks of documented behavior during pull-request workflows.
Documented uses
- Turn PRDs, ADRs, READMEs or documentation sites into a curated requirements corpus.
- Connect selected documentation sources to a repository and derive claims and user flows.
- Generate and run tests against product interfaces to identify documentation-product disagreements.
- Review flow status, coverage and background job progress.
Requirements & limits
TrueCourse supports single-machine local mode without sign-in and team deployment with sign-in. It stores durable data in Postgres. The project is open source under MIT, except for identified enterprise features: document Connections, additional repository providers and multiple workspaces.
- Scanning, flow setup and flow generation require a configured model provider; Flow runs are deterministic and require no model.
- GitHub is the only documented repository connection provider. Local folders are available in local mode; GitLab and Azure DevOps are listed as coming soon.
- A self-hosted instance can use the operator’s Claude Code login for every workspace instead of a workspace-selected provider.
- More than one workspace and additional repository providers are identified as enterprise features.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Set up a workspace and generate requirement flows
Add documentation context, connect a repository, and generate flows for later testing.
Before you start
- A running TrueCourse instance.
- A documentation source: repository markdown or a documentation site with llms.txt.
- A repository connection; GitHub is the documented provider, with local folders available in local mode.
- A model provider configured in workspace Settings for document scanning, flow setup and flow generation.
Steps
In Context, add a repository markdown source or a documentation site through its llms.txt.
Run a Document scan to curate the source documents into a corpus.
In Code, connect a repository and select the sources it should read.
Allow background Flow setup and Flow generation to derive claims, group them into flows and write tests.
Expected outcome
The workspace has generated flows and tests ready for a Flow run.
Guide limitations
- This guide covers setup and generation; it does not document a successful run or deployment procedure.
- GitHub is the only documented repository connection provider; local folders require local mode.
- The documented model providers are Anthropic API, OpenAI, AWS Bedrock, GitHub Copilot and Google Gemini; self-hosted instances can use the operator’s Claude Code login.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 7 November 2026
- Original website title
- TrueCourse · AI writes the code. You own what it does. · Official homepage title · retained in its source language