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 skills add alexex1993/mcu-skills --skill rp2350a-weactgit clone --depth 1 https://github.com/alexex1993/mcu-skillsWrote 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/skills/alexex1993/mcu-skills/rp2350a-weact)<a href="https://agentmods.dev/skills/alexex1993/mcu-skills/rp2350a-weact"><img src="https://agentmods.dev/badge/skills/alexex1993/mcu-skills/rp2350a-weact/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/alexex1993/mcu-skills/rp2350a-weact"><img src="https://agentmods.dev/badge/skills/alexex1993/mcu-skills/rp2350a-weact.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 2 findings, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium analysis-evasion · line 1 Suspicious Unicode normalization or mixed-script contentFix: Review the flagged content for security risks. Ensure no credentials, secrets, or sensitive data are exposed.
- medium Rogue Agent · line 102 Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.Fix: Remove any persistence mechanisms (cron jobs, startup scripts, state files). Skills should not maintain state across sessions without explicit user consent.
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 5.1 | $0.00000 | $0.04723 |
| Opus 5 | $0.00000 | $0.02361 |
| Sonnet 5 | $0.00000 | $0.00945 |
| Haiku 4.5 | $0.00000 | $0.00472 |
Grade A, and why
rp2350a-weact 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 12d 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 — 239 lines — stays where its author put it; the contents beside it link to each section on GitHub.
WeAct Studio RP2350A Core Board (RP2350A)
A Pico-2-shaped RP2350A core board that ships in two incompatible revisions under one name. V2.0 is a pin-for-pin Pico 2 replacement; V1.0 brings out every GPIO instead, and pays for it by moving the power pin, the ADC reference and three GPIO functions. Firmware built for the wrong one compiles cleanly and then drives the wrong pins — so establish the revision before writing any code, and read the reference files rather than assuming Pico 2 behaviour.
reference/board-hardware.md— the complete board reference: how to tell the revisions apart, the 40-pin map for both, the off-header GPIO, power trees, clocks, memory map, vendor-material inventory plus a development guide (Part II: §8 toolchain, §10 peripheral cookbook, §11 RP2350 gotchas including erratum E9 and the PWM slice map, §12 flashing, §13 a symptom → cause → fix table).reference/recipes.md— copy-paste code:platformio.iniwith all four envs, the revision guard, reading the real flash size out of the chip, LEDs, the KEY button, V2.0 power sensing, ADC,Wire1remapping, PWM, EEPROM, LittleFS, core 1.template/— a project that builds clean for both revisions and both flash sizes, in two variants, plus a scaffold script. Seetemplate/README.md.
Orientation
| MCU | RP2350A, QFN-60 — 2 × Cortex-M33 @ 150 MHz (or 2 × Hazard3 RISC-V, not reachable from this toolchain), 30 GPIO |
| Memory | 520 KB SRAM, 8 KB OTP, no internal flash — 4 MB or 16 MB QSPI, XIP from 0x10000000. Max sketch 4,190,208 B / 16,773,120 B (last 4 KB = emulated EEPROM) |
| Clock | 12 MHz crystal → 150 MHz. No 32.768 kHz crystal fitted |
| LED | V1.0: GP25 green + GP24 blue · V2.0: GP25 green only. All active-HIGH through 5.1 kΩ — deliberately dim, and not on the header |
| Button | RESET (→ RUN, not readable) and BOOT (→ QSPI_SS, readable at runtime as if (BOOTSEL)) on both · V1.0 only: 23@KEY on GP23, active-LOW, 5.1 kΩ pull-up |
| USB | USB-C, FS device/host. Serial = USB CDC; 1200-baud touch reboots into BOOTSEL; enumerates 2E8A:000F, BOOTSEL drive is named RP2350 (not RPI-RP2) |
| ADC | 12-bit: A0=GP26, A1=GP27, A2=GP28, A3=GP29 — free input on V1.0, VSYS sense on V2.0 + channel 4 = die temp. analogRead is 10-bit by default. ADC_VREF on header pin 35 (V2.0) or test point T1 only (V1.0) |
| PWM | analogWrite on any GPIO, 8-bit @ 1 kHz. 12 slices exist but only slices 0–7 are reachable: 4 GPIOs per slice, and GPn / GPn+16 share a channel |
| Bus defaults | Wire = 4/5 · Wire1 = 26/27 (= A0/A1!) · SPI = 16/19/18/17 · SPI1 = 12/15/14/13 · Serial1 = 0/1 · Serial2 = 8/9 |
| Power | V1.0: pin 39 5V → LDO, 3.6–6.5 V → 3.3 V @ 800 mA · V2.0: pin 39 VSYS → buck-boost, 1.8–5.5 V. Pin 40 is silkscreened VIN on both and is USB VBUS, not an input |
| Debug | 4 pads at the far end: 3V3 / DIO / CLK / GND. Boot ROM is mask ROM — unbrickable |
| Size | 51.00 × 21.00 mm, rows 17.78 mm apart — same footprint as a Pico 2 |
| Toolchain | PlatformIO + raspberrypi 1.20.0 + arduino-pico 6.0.0 (board = rpipico2, board_build.core = earlephilhower) |
What ships with it
10 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- assets/IMAGE.md 449 B
- reference/board-hardware.md 22 KB
- reference/recipes.md 9.6 KB
- template/.gitignore 13 B
- template/include/board.h 5.7 KB
- template/platformio.ini 2.9 KB
- template/README.md 4.4 KB
- template/src/main.cpp 5.7 KB
- template/variants/minimal/main.cpp 810 B
- template/variants/new-project.sh 2.1 KB runs code
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.
- 12d ago First seen · 239 lines · 0 tokens per session scan A 5c0c92952d47
rp2350a-weact is a skill published in the GitHub repository alexex1993/mcu-skills (17 stars, last pushed 2d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 4,723 tokens. 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 2026-08-30.
Other skills, from other repositories
iot-developer
Expert in IoT development, microcontrollers, sensors, and MQTT protocols.
edge-iot
Edge computing, IoT protocols, and embedded systems integration.
ruview-configure
Configure RuView — ESP32 sdkconfig variants, NVS provisioning, WiFi channel / MAC filter overrides (ADR-060), edge intelligence modules (ADR-041), sensing-server flags, multi-node mesh, and Cognitum Seed integration. Use when adjusting how a deployed RuView system behaves without changing code.
ruview-applications
Run RuView sensing applications — presence/occupancy, breathing & heart rate, activity & fall detection, 17-keypoint pose estimation (WiFlow), sleep monitoring & apnea screening, environment mapping, Mass Casualty Assessment (MAT), and the 3D point-cloud fusion demo. Use when someone wants to actually do something…
ruview-mmwave
Set up and run RuView mmWave / FMCW radar sensing — ESP32-C6 + Seeed MR60BHA2 (60 GHz, heart rate / breathing rate / presence) and HLK-LD2410 (24 GHz, presence + distance), plus mmWave↔WiFi-CSI sensor fusion (48-byte fused vitals, MR60BHA2/LD2410 auto-detect, v0.5.0+). Use when the deployment includes a…
provision-node
Build, flash, and provision an ESP32-S3/C6 CSI node for RuView — firmware variant choice, ESP-IDF Windows-subprocess flow, NVS/WiFi/channel/MAC-filter overrides.