How do I install the Tuyaopen/onchip Watchdog skill?

Run npx skillmds@latest add tuya/tuyaopen-onchip-watchdog in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Tuyaopen/onchip Watchdog safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Tuyaopen/onchip Watchdog?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Tuyaopen/onchip Watchdog free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Tuyaopen/onchip Watchdog?

tuya (@tuya) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Tuyaopen/onchip Watchdog by tuya — Agent Skill

Name: Tuyaopen/onchip Watchdog
Author: tuya

Tuyaopen/onchip Watchdog

On-chip watchdog timer for TuyaOpen using the tkl_watchdog API: init with a timeout, then refresh ("feed") periodically so the MCU doesn't reset. 看门狗、watchdog、喂狗、复位保护、超时复位、tkl_watchdog、防死机。

tuya Updated Sep 17, 2026

File contents

TuyaOpen On-Chip Watchdog

On-chip — no Kconfig, no TDD

Platform-provided. Do NOT write CONFIG_ENABLE_WATCHDOG. No app_default.config change.
No board_register_hardware(). Call tkl_watchdog_* directly. There is a single system watchdog (no instance id / pins) — record as onchip:watchdog in used-peripherals.json if you want it on the diagram.

Init + feed

tkl_watchdog_init() returns the actual refresh interval the hardware granted (may differ from what you asked). Refresh before that interval elapses, from a point that only runs when the firmware is healthy.

#include "tal_api.h"
#include "tkl_watchdog.h"

#define WDT_TIMEOUT_MS   10000

void app_watchdog_start(void)
{
    TUYA_WDOG_BASE_CFG_T cfg = { .interval_ms = WDT_TIMEOUT_MS };
    uint32_t actual_ms = tkl_watchdog_init(&cfg);   /* hardware-granted interval */
    PR_NOTICE("watchdog interval = %u ms", actual_ms);
}

/* Call this well within the interval, on your healthy heartbeat (e.g. main loop). */
void app_watchdog_feed(void)
{
    tkl_watchdog_refresh();
}

API Reference

Function	Description
`tkl_watchdog_init(&cfg)`	`cfg.interval_ms` = requested timeout. Returns the actual granted interval (ms)
`tkl_watchdog_refresh()`	Feed the dog — call before the interval expires
`tkl_watchdog_deinit()`	Disable the watchdog

Troubleshooting

Symptom	Likely cause	Fix
Device keeps resetting	Not feeding in time	Refresh well within the returned interval, not the requested one
Watchdog never resets a hang	Feeding from a timer/ISR that runs regardless of health	Feed only from a path that stops when the app hangs (main loop)
Reset during long blocking op	Interval shorter than the operation	Increase `interval_ms`, or feed inside the long op at safe points

Reference example

SDK: examples/peripherals/watchdog/ (init with timeout, periodic refresh).

---name: tuyaopen-onchip-watchdogdescription: On-chip watchdog timer for TuyaOpen using the tkl_watchdog API: init with a timeout, then refresh ("feed") periodically so the MCU doesn't reset. 看门狗、watchdog、喂狗、复位保护、超时复位、tkl_watchdog、防死机。---# TuyaOpen On-Chip Watchdog## On-chip — no Kconfig, no TDD- Platform-provided. **Do NOT** write `CONFIG_ENABLE_WATCHDOG`. No `app_default.config` change.- No `board_register_hardware()`. Call `tkl_watchdog_*` directly. There is a single system watchdog (no instance id / pins) — record as `onchip:watchdog` in `used-peripherals.json` if you want it on the diagram.## Init + feed`tkl_watchdog_init()` returns the **actual** refresh interval the hardware granted(may differ from what you asked). Refresh **before** that interval elapses, from apoint that only runs when the firmware is healthy.```c#include "tal_api.h"#include "tkl_watchdog.h"#define WDT_TIMEOUT_MS 10000void app_watchdog_start(void){ TUYA_WDOG_BASE_CFG_T cfg = { .interval_ms = WDT_TIMEOUT_MS }; uint32_t actual_ms = tkl_watchdog_init(&cfg); /* hardware-granted interval */ PR_NOTICE("watchdog interval = %u ms", actual_ms);}/* Call this well within the interval, on your healthy heartbeat (e.g. main loop). */void app_watchdog_feed(void){ tkl_watchdog_refresh();}```## API Reference| Function | Description ||----------|-------------|| `tkl_watchdog_init(&cfg)` | `cfg.interval_ms` = requested timeout. **Returns** the actual granted interval (ms) || `tkl_watchdog_refresh()` | Feed the dog — call before the interval expires || `tkl_watchdog_deinit()` | Disable the watchdog |## Troubleshooting| Symptom | Likely cause | Fix ||---------|--------------|-----|| Device keeps resetting | Not feeding in time | Refresh well within the **returned** interval, not the requested one || Watchdog never resets a hang | Feeding from a timer/ISR that runs regardless of health | Feed only from a path that stops when the app hangs (main loop) || Reset during long blocking op | Interval shorter than the operation | Increase `interval_ms`, or feed inside the long op at safe points |## Reference exampleSDK: `examples/peripherals/watchdog/` (init with timeout, periodic refresh).

Frequently asked questions