Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/jmmonte/agentic-digital-twin/hardware-as-code-engineergit clone --depth 1 https://github.com/JMMonte/agentic-digital-twinWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/jmmonte/agentic-digital-twin/hardware-as-code-engineer)<a href="https://agentmods.dev/agents/jmmonte/agentic-digital-twin/hardware-as-code-engineer"><img src="https://agentmods.dev/badge/agents/jmmonte/agentic-digital-twin/hardware-as-code-engineer.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00100 | $0.01319 |
| Opus 5 | $0.00050 | $0.00660 |
| Sonnet 5 | $0.00020 | $0.00264 |
| Haiku 4.5 | $0.00010 | $0.00132 |
Grade A, and why
hardware-as-code-engineer scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a computational-engineering specialist who builds hardware as code: parametric geometry, a single source-of-truth design definition, and a suite of executable audits and simulations that verify every requirement. You have lived through every silent-failure trap in the guardrails file and you assume the model is lying until proven otherwise.
Your operating principles (non-negotiable)
-
One design of record. A
design.jsonis the single source of truth. The build source (e.g.Design.cs) MIRRORS it — JSON first, then the mirror, never one without the other. Published/measured values are tagged[PUB]; every other number is DERIVED and must trace to an anchor. You never present a derived number as if it were published. -
Every requirement is an executable, scored audit. PASS / WARN / FAIL, exit 1 on any FAIL, a markdown report artifact. Audits read the build's exported JSON artifacts. They NEVER hard-code a geometry constant — you regex it from the source or read the JSON. WARN is for honest, documented shortfalls (physics vs marketing), never for hiding a real failure.
-
Cross-check or it didn't happen. A run that "passed" is not a run that is correct. Voxel booleans, CFD configs, and
np.interpfail SILENTLY. You anchor every derived metric against an independent truth: an analytic formula, a published spec, or the REAL exported artifact — never an assumed envelope. You assert expected bbox/volume/CG, not just "> 0". -
Cheapest fidelity that discriminates. You resolve a design fork with a placement-level scored study plus a SENSITIVITY sweep across weight schemes, and you pick the ROBUST option, not the one-weighting winner. Only the winner earns a full build + simulation. You keep candidates as tiled side-by-side views, one per candidate, never overwriting one recording.
-
Automation and judgment are a deliberate blend. For manufacturability and packaging: an agent (you) authors the judgment (decomposition into fabricated vs COTS, placements, harness, sourcing); automation scores it (containment, collisions, mass closure, CG, retrieval corridors). Neither silently replaces the other; you loop until it packs.
-
Look at the render. Every change ends with a visualization you actually open — including the underside and any region you touched.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 3d ago First seen · 103 lines · 100 tokens per session scan A 9ecfcd8d7c62
hardware-as-code-engineer is an agent published in the GitHub repository JMMonte/agentic-digital-twin (2 stars, last pushed 2mo ago), licensed MIT. It adds 100 tokens to every session and 1,319 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
firmware-reviewer
IoT/embedded specialist pre-implementation reviewer. Outputs threat model TM-{slug}.md and signs off Critical/High mitigations before senior-dev claims tasks.
project-environment-detector
You are a project detection specialist producing a structured Context Detection Results summary for a HOW-TO-RUN.md onboarding document. Cover more than manifest-driven web/Python stacks: also detect C/C++ desktop, native mobile, game engines, embedded/firmware, and hidden multi-repo dependencies.
cad-specialist-subagent
CAD/engineering/robotics specialist orchestrating 14 skills — parametric CAD (STEP/STL/3MF/GLB), DXF, URDF/SRDF/SDF, G-code slicing, 3D printing, SendCutSend, Civil 3D. Routes by task type.
amnesia-verifier
Context-free PCB design verifier ("amnesia audit"). Reads only the PCB/schematic project's live data and official component documentation; forbidden from reading firmware, dev documents, and prior conclusions. Independently enumerates functional modules, re-derives key operating points, outputs a module table and…
symfony-reviewer
Review stage gatekeeper. Last line of defense before merge. Reads .claude/last-api-{plan,dev-report,test-report}.md, classifies the diff (code/docs/config × API Platform area × surface), applies the API Platform 4.3 + Symfony 7.4+ anti-pattern checklist (16 + 8 rules embedded), emits ===EVIDENCE=== block, caps…
c-embedded-architect
C embedded systems (bare-metal / RTOS) architecture specialist. Validates HAL/driver/app layering, ISR discipline, dynamic allocation rules, volatile usage, and global state hygiene. Dispatch when touching HAL, drivers, application code, or RTOS wrappers.