StudyLoop is an early open-source release. This roadmap describes product outcomes rather than promising dates or version numbers.
The current release supports:
- local Study Session and Body Double workspaces with supported AI CLI agents;
- energy-aware Socratic sessions, parking, timers, and session close-out;
- flashcards, quizzes, spaced repetition, and a single recommended next action;
- local content generation from Markdown, text, and supported PDF sources;
- study plans stored as readable Markdown with session checkpoints;
- searchable session history and optional Obsidian export;
- laptop and tablet Web UI layouts;
- a choice of second brain (shipped in 0.3.0):
studyloop brain destination setand the Today panel's launcher let a plan and today's next action appear where a learner already keeps their thinking. Obsidian is the free, local option and works now; support for a connected visual planner is served through the learner's own assistant rather than by StudyLoop holding a credential.
Work in this phase is about trustworthy installation, recovery, accessibility, and documentation. A feature is not considered ready merely because a backend route or design document exists; the user-facing path and its evidence must work.
The session-memory reliability candidate focuses on recovering local coding conversations, repeatable repair and configured sync. Release acceptance remains pending. Its guide states the current boundaries.
Richer shared memory remains separate research: enforced work/personal scope, forgetting across sync and restore, source-grounded decision explanations and review of conflicting advice. Those capabilities are not promises of the reliability release, and no database replacement is required for that release.
The next product improvements are:
- a simpler installation and upgrade path than a source checkout;
- continuity from an active plan into the next session: the plan already
biases
studyloop nowand Today with tested ranking rules and the architect interview runs from the CLI, the Web UI and over MCP, but a live study session is still not bound to a plan and checkpoints and milestones are never inferred from session events (see the Study Plans guide, "Deliberately not automatic"); - clearer in-product explanations when an agent, voice backend, or optional integration is unavailable;
- broader manual accessibility testing and more contributor-friendly UI evidence.
These are useful directions, but they should not weaken the current local-first workflow:
- a supported phone experience designed for the smaller screen rather than a compressed desktop layout;
- offline review where the data and conflict model are explicit;
- easier community-created courses and learning templates;
- localisation;
- additional agent and local-model integrations with the same session guarantees.
StudyLoop does not currently promise a hosted service, mobile app, PyPI package, Homebrew formula, phone support, or offline Web UI. Historical implementation plans live in the repository for maintainers, but they are deliberately excluded from the public documentation site.
Have a workflow that should influence the order? Open an issue with the learning problem, the device or agent involved, and what a successful outcome would look like.