Code defects: truecourse analyze
From the categories linters cover (unused code, style, missing types) through to ones they don’t reach: circular dependencies, layer violations, dead modules, race conditions, security anti-patterns, performance footguns. Tree-sitter analysis combined with LLM review: 1,500+ deterministic rules and 100 LLM rules across JavaScript, TypeScript, Python, and C#.
Business-logic drift: truecourse guard
Catches the implementation no longer matching what the docs say it should do. TrueCourse curates your PRDs, ADRs, and READMEs into a spec corpus, an LLM authors scenario tests bound to each spec section once, and
guard run executes them deterministically. A failing scenario means that section and the code disagree..truecourse/ in your repo and surface them in a shared dashboard for human review, with plain-text CLI output an agent can read directly.
No setup, no database
TrueCourse creates.truecourse/ in your repo on first use and stores everything there as plain JSON files; there is nothing to provision. For LLM-powered work it uses the Claude Code CLI by default, or a provider API with your own key (Anthropic, OpenAI, AWS Bedrock, GitHub Copilot). Your first truecourse command asks which, and truecourse config llm setup changes it later. With neither available, deterministic analysis still runs and LLM-dependent features are skipped.
Get started
Quickstart
Install TrueCourse and run your first analysis and guard cycle in minutes.
Installation
Prerequisites, the npm package, and the first-run LLM choice.
Dashboard
One web UI for both capabilities: the architecture graph, violations, analytics, and the spec-curation + guard workflow.
CLI reference
Every command and flag in one place.