FlashLearnStart with what exists.
Bring together the knowledge already living in your repositories, documentation, and team wikis.
Today: local repositories, Markdown, JavaScript/TypeScript, and Go.
Planned: wiki and other knowledge-source connectors.
Agentic AI × learning velocity
Turn project knowledge
into everyday momentum.
Onboard to unfamiliar projects faster. Source what your team already knows, generate useful context, and build lasting familiarity through regular, gamified spaced repetition.
Open-source · local-first · MIT licensed
Three stages. One learning loop.
From scattered information to confident contribution.
Make the context actionable.
Generate focused, contextualized knowledge snippets that explain how a project works and build familiarity with its concepts, decisions, and code.
Today: source-attributed cards, quality filtering, and optional AI generation through a configured endpoint.
Build understanding every day.
Learn through regular, gamified spaced repetition woven into your daily workflow—with a path to Teams, GitHub, and Slack integrations.
Today: a local Teams-style review UI and spaced-repetition engine.
Planned: native Teams, GitHub, and Slack delivery integrations.
Your project. Your learning loop.
Cards and canonical spaced-repetition review state live in your project's .flashlearn/ directory. The live review client keeps topic-insights event history in browser localStorage on that device; it is not written to .flashlearn/ or synchronized across browsers or devices, and clearing site data removes it. Offline extraction stays local; enabling an AI endpoint sends selected code to that endpoint, subject to its permissions and retention policies.
The public demo contains only curated public samples. It never accesses your repository and never submits ratings to a backend.
Read the data boundaries →