# Agent2Learn

> A local course vault for your own Waterloo LEARN account. Original files, Markdown twins, and inspectable source citations. No hosted course storage or product telemetry. Grades and discussions are off by default. Uploads are disabled in this build. The evidence scan does not establish correctness or policy compliance.

PyPI release: 0.1.2.

## Start here

- [Setup prompt](/agent-prompt.txt): instructions an agent can use to prepare setup and hand interactive onboarding to the user.

## Documentation

- [Authentication](/docs/authentication/): Sign in on your own device and recover an expired LEARN session.

- [Command reference](/docs/commands/): The complete v0.1 command surface, grouped by what you want to do.

- [Scan a draft](/docs/evidence-scan/): Understand the experimental lexical evidence scan and its limits.

- [For agents](/docs/for-agents/): A copyable setup prompt and a source-first workflow for your coding agent.

- [Installation](/docs/installation/): Install the engine, then create your first local course vault.

- [A course vault your agent can read.](/docs/introduction/): What Agent2Learn does, and where to start.

- [Privacy](/docs/privacy/): What stays local, what is collected, and what uses the network.

- [Study with sources](/docs/study/): Find course material, assemble a grounding pack, and follow citations.

- [Your first sync](/docs/sync/): Bring courses local, choose download scope, and understand coverage.

- [Troubleshooting](/docs/troubleshooting/): One useful next step for common installation, authentication, and coverage problems.

- [The vault](/docs/vault/): An ordinary folder, with originals, Markdown twins, indexes, and revision history.

## Canonical sources

- [Repository](https://github.com/ManagementMO/agent2learn): Python engine, source documentation, release gates, and canonical skills.

- [Agent Skills](https://github.com/ManagementMO/agent2learn/tree/main/skills): a2l-setup, a2l-sync, a2l-study, and a2l-coursework.