Method · repair bench
AI on the repair bench, grounded in your documents.
Drop a schematic PDF and a boardview into WrenchBoard and, about two minutes later, you are working on a device the system actually knows: components inventoried, nets classified, boot sequence mapped, failure rules audited. Here is what happens in those two minutes, and why nothing the agent tells you afterwards is left to chance.
Open a board on the bench01Two minutes
Two minutes from documents to working knowledge
A senior technician who has seen a board a hundred times carries its architecture in their head. WrenchBoard builds that mental model explicitly, before the first probe touches the board.
The build is done by a small factory of specialized Claude roles: a Scout gathers what is known about the device, a Registry freezes the component inventory, three Writers work in parallel (one maps the board, one writes the clinical failure picture, one compiles the vocabulary), and an Auditor checks the result before anything is stored. From a schematic PDF and a boardview, the whole pass takes about two minutes.
The output is not prose, it is a queryable corpus on disk, per device. Every diagnostic session that follows reads from it and writes back to it: measurements, confirmed causes, open questions. Nothing restarts from zero.
02The graph
The schematic becomes a graph you can query
A sixty-page schematic PDF is documentation for humans with hours to spend. WrenchBoard compiles it, page by page, using Claude vision, into an electrical graph: which nets exist, what class they belong to, which components sit on them, what the boot sequence looks like, with a quality report attached stating how much of the document was confidently read.
That compilation is what turns “search the PDF” into “ask the board a question”. What a specific component does, what shares its rail, what should be live at this boot phase: answers come from the graph in seconds, and they cite the graph, not the model’s memory. The compiler itself has its own page.
03The engines
Deterministic engines under the reasoning
The agent does not guess causal chains, it delegates them. Two pure algorithms sit at the core of the diagnostic stack and make zero language-model calls at runtime:
- The simulator advances the board’s boot sequence phase by phase and emits a timeline: which rails are dead, which components are unpowered, what blocks each phase.
- The hypothesizer takes a partial observation, the handful of measurements you actually made, and enumerates the one- and two-fault candidates that would explain it, ranked by how well each one fits.
Deterministic means checkable: the same measurements always produce the same candidates, and every causal chain can be verified by hand against the graph. The agent’s job is to reason on top of that scaffolding in plain language, at the level of the technician using it, not to improvise electronics.
04The session
A session at the bench, iron in hand
At the bench, the agent drives the boardview while you work: it highlights the pin to probe, traces the net under suspicion, annotates the measurement protocol on the board itself. If its reasoning needs eyes, it requests a frame from your USB microscope; you can also drop a close-up of a suspect chip into the chat.
You measure, it interprets, the loop continues. Every measurement is logged, every capture is persisted, and the whole session can be replayed afterwards: what was said, what was decided, what the agent actually saw. The full diagnostic workflow is structured the way a senior would run it, but the iron never leaves your hand.
The reading rides along with your next message
BAT1FUSED · 0 V · 14:02 You do not retype it, and the agent reads it before it answers.
05The limits
The honest limits
WrenchBoard is at its strongest with a schematic and a boardview in hand; that is where verified inventories and deterministic chains operate at full capacity. Without those documents, it still builds a first knowledge pack from public sources: useful, but thinner. That gap is not a software limitation, it is a documentation-access problem, and it is the real subject behind “too complex to repair”. The canonical development target is the MNT Reform mainboard, chosen precisely because its documentation is open under the CERN-OHL-S license.
Watch it work
The same agent, on a real board
This is not a mockup, it is the bench itself with its own interface, running on the MNT Reform pack exported from the engine: 487 real components and 2066 real pads. Every step below is what the engine answers on this board, from the device's rules for the symptom to the schematic graph behind the input fuse F1 and the board net it feeds. Here is what the agent does with a board that will not power up.
Eight tool calls, each one a real answer from the engine on this board, and not one component named from memory.
FAQ
Frequently asked questions
What exactly is a knowledge pack?
A structured, per-device corpus the system builds before any diagnosis: component inventory, classified nets, inferred boot sequence, known failure rules. It is assembled by four specialized Claude roles (a Scout, a Registry, three parallel Writers, an Auditor) in about two minutes, then stored on disk where every later session can query and enrich it.
Do I need the schematic and the boardview of my device?
The system performs at its best with both: that is where the deterministic engines produce verifiable causal chains. Without them it does not collapse; it builds a first knowledge pack from public sources. The gap between the two is exactly the gap right-to-repair laws are meant to close.
What does deterministic mean here, concretely?
Two engines sit under the agent's reasoning: a boot-sequence simulator that advances phase by phase and emits a timeline of dead rails and dead components, and a hypothesizer that enumerates one- and two-fault candidates explaining a partial observation, ranked against what you measured. Neither makes a language-model call at runtime: same input, same output, every chain checkable by hand.
Can the agent actually see my board?
Yes, on your cue. Plug a USB microscope or webcam into the bench and the agent requests a frame when its reasoning needs one; you can also drop a macro shot into the chat at any time. Captures are persisted with the repair, so a session can be replayed end to end: words, decisions, and the photographs the agent looked at.
What happens if the agent cites a component that does not exist?
It cannot assert one. A reference lookup that fails returns the closest real matches, and the agent must pick from those or ask you. As a second layer, a server-side sanitizer scans every outgoing message and visibly marks any reference it cannot verify against the parsed board before the text reaches your screen.
Get started
Try Wrench Board free, in your browser.
Free plan, no card required. Open the app, pick a covered device, and start a diagnostic session with the agent.