Demo
Write. Check. Fix. Green.
Draft on a virtual board. Fail for a real reason. Patch. Pass.
-
01
Write
Describe the board and the task.
-
02
Check
Runs on a virtual board.
-
03
Fix
Wrong output. Patch and re-run.
-
04
Green
Only when behavior matches.
Connect
Your agent, or ours.
Same virtual board either way. Physical boards use popular probes (not OpenOCD configs).
LabWired Agent
CLI + skills. Local model optional.
curl -fsSL https://labwired.com/agent-install.sh | sh
Claude or Codex
MCP into the same tools.
claude mcp add labwired --transport http https://api.labwired.com/mcp
Skills
Included.
verify-firmwareRun the check
diagnose-firmwareFail, patch, re-check
inspect-evidenceRead a result
board-bringupBoard and wiring
scaffold-firmwareBlink / serial hello
report-evidenceReport for CI
FAQ
Questions.
Do I need a board?
No. Checks run on a virtual board. Use hardware when you want.
Does the model grade itself?
No. The virtual board has to match. Compile success is not enough.
Claude or Codex?
Yes. MCP uses the same tools as the CLI agent.