Overview¶
Hydron is a hardware-native AI coding assistant for embedded and system software engineers. It provides spec-aware code generation grounded in a Hardware Knowledge Graph, context-indexed understanding of codebases, datasheets, schematics, and BSPs, and hardware-in-the-loop functionality.
Two surfaces, one engine¶
- VS Code extension. Inline diffs, code lenses, and a graphical serial monitor inside your editor.
- CLI. Headless and SSH setups, scripting workflows, and terminal-first development.
Both share the same context engine, agent modes, and hardware integrations.
What grounds the answers¶
| Source | What it gives you |
|---|---|
| Hardware Registry | Over 500 pre-indexed components, with vendor datasheets, reference manuals, and errata |
| Datasources | Your own codebases, schematics, BSPs, and proprietary specifications |
| Project Dashboard | The project that ties them together and syncs to every client |
Without a project selected, Hydron answers from training data. With one, it answers from your stack.
What it does¶
- Spec-aware code generation against your actual register maps rather than a reference design
- Agent modes for exploring, planning, debugging, and implementing
- Hardware in the loop through a serial monitor the agent can drive
- Plugins for vendor toolchains, bring-up routines, and internal standards
For teams¶
Projects carry their context to everyone on them. A new engineer inherits the same catalogue, files, and indexed codebase on joining. Roles control who can change what the team's answers are grounded in.
Enterprise plans, DPAs, and SLAs are available. See Teams and roles.
Start here¶
- Quickstart to authenticate and run a first prompt.
- Project Dashboard to give Hydron your hardware context.
- Best practices for the patterns that make answers accurate.
Support¶
Email support@h2loop.ai.