agentleFS
wedsamuel1230wedsamuel1230/embedded-engineering-skills/skills/arduino-workflow-router/SKILL.md

arduino-workflow-router

Skill28 starsChanged 31 days agoMITView on GitHub

Route Arduino and embedded-system requests across Arduino IDE, Arduino CLI, PlatformIO, and vendor-specific tools. Use when a request combines firmware, board constraints, electronics, power, networking, calibration, enclosure, upload, debugging, OTA, or maintenance, or when the exact board or toolchain is not yet clear.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

12 rules. About 0 are common, 3 are distinctive, and 9 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 9

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill as the concise entrypoint for complete or cross-discipline embedded work.

    line 9
  2. Very specific:

    If the work spans sessions or physical actions, embedded-project-loop is the recommended first skill.

    line 22
  3. Very specific:

    When a named board or board family must be resolved, load ../board-support/SKILL.md and its indexed profile before selecting pins.

    line 27
  4. Very specific:

    Read references/board-intake.md for board and hardware checks.

    line 29
  5. Very specific:

    Read references/toolchain-selection.md for IDE, CLI, PlatformIO, or vendor version and library compatibility.

    line 30
  6. Very specific:

    Read references/failure-recovery.md before upload, boot, power, or firmware recovery actions.

    line 33
  7. Very specific:

    Do not load both arduino-serial-monitor and a second serial-debugging skill.

    line 56
  8. Distinctive:

    Do not invent a pin, voltage, current limit, memory size, peripheral, protocol, library version, or upload command.

    line 61
  9. Very specific:

    Default combined order: arduino-workflow-router -> board-selection -> board-support -> pin-assignment -> wiring-safety-check -> …

    line 120
  10. Very specific:

    If the user already supplied an exact, supported board identity, skip board-selection and begin with board-support.

    line 127
  11. Distinctive:

    Do not treat a table row as proof that a skill was run.

    line 146
  12. Very specific:

    Use ../../docs/arduino-skill-contract.md: state assumptions, required tools and versions, implementation steps, tests/evidence by proof …

    line 152

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Embedded Workflow Router

Use this skill as the concise entrypoint for complete or cross-discipline embedded work. For a focused task with a confirmed board and one stage, load the specialist skill directly.

Recommended lifecycle entry

For any physical or multi-session request, load embedded-project-loop first. It establishes the goal, one next action, evidence boundary, rollback path, and user-owned physical gate. Then return here for board and toolchain routing.

Load First

  1. Read ../../docs/arduino-skill-contract.md.
  2. If the work spans sessions or physical actions, embedded-project-loop is the recommended first skill; create its goal/next-todo state before implementation and keep it open until the evidence gate is satisfied.
  3. Fill ../../docs/board-support/board-profile-template.md or state why it is unnecessary.
  4. When a named board or board family must be resolved, load ../board-support/SKILL.md and its indexed profile before selecting pins.
  5. Read references/board-intake.md for board and hardware checks.
  6. Read references/toolchain-selection.md for IDE, CLI, PlatformIO, or vendor version and library compatibility.
  7. Select specialist skills from the routing table below.
  8. Read references/failure-recovery.md before upload, boot, power, or firmware recovery actions; read references/connected-device-security.md for any networked or updateable device.

This is the load-first order. Load on demand: references are loaded only when their trigger applies, so the router stays under 500 lines and detailed variants remain progressive disclosure.

Trigger precedence

  • For combined, underspecified, or cross-discipline requests, this router owns the first route. Load it before any overlapping specialist, including arduino-project-builder or board-selection.
  • For physical or multi-session requests, load embedded-project-loop before this router, then return to this router for the combined specialist route.
  • For a single-stage request with a confirmed board and toolchain, load the narrowest specialist directly.
  • For a named-board reference or capability lookup, board-support owns the exact identity and source-backed profile. board-selection owns choosing or replacing a board from requirements and may consume its handoff.
  • A board-choice request without implementation, wiring, or lifecycle work may start at board-selection; once another discipline appears, return to this router and preserve the combined-workflow order.
  • Do not load both arduino-serial-monitor and a second serial-debugging skill; the existing serial monitor owns structured runtime evidence.

Intake Gate

Do not invent a pin, voltage, current limit, memory size, peripheral, protocol, library version, or upload command. Ask for the missing value or proceed with a clearly labeled assumption and a verification step. Record the exact board, framework, toolchain, host, dependency versions, and desired proof stage.

Routing Table

NeedLoad nextEvidence focus
Zephyr Developmentzephyr-developmenttarget-specific evidence
STM32 Developmentstm32-developmenttarget-specific evidence
Pico SDK Developmentpico-sdk-developmenttarget-specific evidence
MicroPython Developmentmicropython-developmenttarget-specific evidence
Probe Debuggingprobe-debuggingtarget-specific evidence
Fault Forensicsfault-forensicstarget-specific evidence
Bus Debuggingbus-debuggingtarget-specific evidence
Boot and Flash Recoveryboot-flash-recoverytarget-specific evidence
Memory and Timing Debuggingmemory-timing-debuggingtarget-specific evidence
USB Debuggingusb-debuggingtarget-specific evidence
Requirements or board choiceboard-selection, then board-support, then arduino-project-builderdecision and design
Named board reference or capability lookupboard-support, then the relevant specialistexact identity and source-backed constraints
Pin map or GPIO declarationsboard-support, then pin-assignment, then wiring-safety-checkboard constraints and hardware
Wiring, voltage, current, or pull-upswiring-safety-check, power-budget-calculator, circuit-debuggerhardware
ADC/sensor noise, sampling, filtering, or signal conditioningboard-support, wiring-safety-check, sensor-signal-filtering, then sensor-calibration-workbench after detection is provensignal-chain design and measured hardware behavior
Code pattern or board abstractionarduino-code-generator, non-blocking-patterns, memory-budgetingbuild and memory
Library or framework dependencylibrary-selection, then the code/project skillcompatibility and memory
Datasheet, component, or protocol uncertaintydatasheet-interpreter, i2c-bringup-diagnosticianhardware assumptions
Compile, board discovery, upload, or portarduino-cli-skill plus error-message-explainerbuild and upload
PlatformIO, native ESP-IDF, mixed builds, CI or SDK migrationembedded-build-workflow; use arduino-cli-skill only for an Arduino CLI branchexplicit framework/manager and build matrix
Control-loop deadline, actuator ownership, overload or action timingrealtime-control-design, then freertos-patterns for primitivesdesign budget and measured timing
UDP/ESP-NOW telemetry, AP+STA coexistence or radio trace analysiswireless-telemetry, then arduino-serial-monitor for capture/desktop integrationfreshness, protocol and measurement scope
Runtime logs or field symptomsarduino-serial-monitor, error-message-explainer, hardware-tddhardware and system
Timing, blocking, or watchdog risknon-blocking-patterns, hardware-tddbuild and system
Memory or footprint riskmemory-budgeting, library-selectionbuild and runtime
Calibration or sensor driftsensor-calibration-workbench after detection is provensystem
Complete applicationarduino-project-builder, then the relevant domain skillsdesign and build
Connected update or field deploymentota-deployment-guardian plus the security referencedeployment
Parts, PCB-adjacent, or enclosure workbom-generator, enclosure-designer, readme-generatordesign and maintenance
Host/simulation/target test planhardware-tdd and embedded-project-loop when physical work is pendingevidence
Multi-session, physical, or recovery workembedded-project-loop first, then this routerdurable state and user-owned gate

Combined Workflows

Toolchain precedence: when a repository selects PlatformIO or a native SDK, embedded-build-workflow owns build intake. Do not route native app_main, Zephyr, Pico SDK or Rust projects to Arduino CLI by default. See the support matrix before claiming a framework, target or host is validated. Consult the active BottleSumo lessons for mixed builds, separate flash/PSRAM settings, wireless evidence, OTA and bench control.

Skills can be used together. Keep one owner for each decision and pass its artifacts to the next skill. For a combined request, use this concise default order:

loop (if long-running) -> board selection/intake -> wiring safety -> library/memory -> project/code/timing -> toolchain build/upload -> serial and hardware tests -> system/calibration -> deployment/security/maintenance

Default combined order: arduino-workflow-router -> board-selection -> board-support -> pin-assignment -> wiring-safety-check -> sensor-signal-filtering -> non-blocking-patterns -> arduino-serial-monitor -> hardware-tdd. Start with embedded-project-loop when the work spans sessions or physical gates, then keep its next-todo and evidence ledger open through the later stages.

If the user already supplied an exact, supported board identity, skip board-selection and begin with board-support; never skip board-support before pin or electrical advice.

  • Battery-powered Wi-Fi sensor: board intake -> datasheet -> power -> project builder -> code generator -> toolchain -> serial/calibration -> OTA security -> maintenance README.
  • Uno R4 WiFi upload incident: board-family reference -> IDE/CLI discovery -> serial and error diagnosis -> USB/power checks -> boot recovery -> upload proof -> runtime/system proof. See ../../docs/board-support/uno-r4-family.md.
  • Robot controller: board intake -> project builder -> power/BOM -> circuit and code -> serial/system tests -> enclosure -> signed update and rollback.
  • Multi-board sensor library: board profiles -> datasheets/protocol bringup -> code generator -> IDE/CLI/PlatformIO branches -> per-board build proof and documented unsupported behavior.
  • Four-button ESP32 controller: board selection -> pin assignment -> wiring safety -> non-blocking debounce -> code generator -> build proof -> hardware gate -> serial/system evidence.

Do not treat a table row as proof that a skill was run. Report which skills were actually used, which artifacts they produced, and which evidence stages remain unverified.

Output Contract

Use ../../docs/arduino-skill-contract.md: state assumptions, required tools and versions, implementation steps, tests/evidence by proof stage, known limitations, and recovery/security notes.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.