BrightInvestor

Open VS Code and code as usual; BrightInvestor handles the busywork. As you write, it reviews the active file and related modules, pointing out fragile logic, edge cases, and style drift. Click a suggested fix to insert a patch, or open the diff to tweak it before applying. Use the Command Palette to ask for alternatives, stricter type hints, or faster loops. Set project rules once, and BrightInvestor follows your conventions automatically—naming, imports, formatting, and common utilities stay consistent without you jumping between tools.

When tests fail or a red underline appears, pull up the chat panel and ask direct questions: “Why is parseDate allocating so much?” or “Convert this callback to async/await.” The assistant reads the current buffer, nearby files, test output, and your latest commit to propose a small, targeted change. You can run the patch in a scratch preview, generate unit tests to confirm behavior, then accept the change. For larger refactors, request a plan: it will list steps, create staged diffs, and annotate each change with reasoning so code review is faster.

For day‑to‑day teamwork, use BrightInvestor as a pre‑commit gate. Trigger a quick scan to catch risky patterns, missing null checks, or performance footguns before you push. Ask it to draft docstrings from the function body, update examples in your README, or produce a clean pull‑request description that summarizes impact and tests. It mirrors your repository’s tone and style, learning from merges you accept and suggestions you dismiss, so it becomes stricter or looser based on your history. Schedule background sweeps to surface hotspots by file and rule, then open guided fixes when you’re ready.

Onboarding and legacy maintenance are simpler too. Open an unfamiliar service and request a map of key types, entry points, and data flows; jump from that outline to the exact files you need. Ask “where should caching live?” or “what breaks if we bump this dependency?”, and get scoped answers with candidate patches. If you’re preparing a release, have it assemble a changelog from commit messages and labels, then generate smoke‑test scripts to run after deploy. The result: less context switching, clearer intent in diffs, and a steady path from idea to reviewed code.