Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/Jinchen-Yang/diansai-skillnpx agentmods add skills/jinchen-yang/diansai-skill/vision-scaffoldWrote 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/jinchen-yang/diansai-skill/vision-scaffold)<a href="https://agentmods.dev/skills/jinchen-yang/diansai-skill/vision-scaffold"><img src="https://agentmods.dev/badge/skills/jinchen-yang/diansai-skill/vision-scaffold/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/jinchen-yang/diansai-skill/vision-scaffold"><img src="https://agentmods.dev/badge/skills/jinchen-yang/diansai-skill/vision-scaffold.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00099 | $0.00782 |
| Opus 5 | $0.00049 | $0.00391 |
| Sonnet 5 | $0.00020 | $0.00156 |
| Haiku 4.5 | $0.00010 | $0.00078 |
Grade A, and why
vision-scaffold 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.
What it actually says
vision-scaffold —— K230 视觉骨架（流水线 ⑧，算法 lane）
lane: 算法（DeepSeek/GLM 擅长代码） · 门: 语法检查 + 现场标阈值
K230 是独立板，可桌面独立开发。发帧只用 contracts/protocol.py（与主控 protocol.h 同源同签名，杜绝漂移）。
前置
design/solution.md（视觉要做什么）、contracts/protocol.py（已生成）。
生成结构（写到 vision/）
vision/
README.md # CanMV 烧录/运行(指向 env/vision.md), 镜像版本锁定提醒
main.py # 主循环: 取流 → 处理 → 用 protocol.py 发帧
line_follow.py # 巡线: 取一行/多行, 求中线偏差 error(int16)
blob_track.py # 色块/光点: find_blobs → cx,cy
digit_recog.py # 门牌数字: KPU/模板, 出 class_id+confidence
protocol.py # 从 ../contracts/ 复制(勿手改)
config.py # ★ 阈值/ROI/曝光 占位, 现场标定
步骤
main.py：CanMV 初始化（sensor/lcd），主循环里按当前任务调 line_follow/blob_track/digit_recog，把结果用protocol.line_error()/blob_xy()/target_class()经 UART 发出。硬件相关 import（sensor/image/lib）放运行时，保证主机py_compile语法可过。line_follow.py：灰度/二值化取中线，输出error（中线相对画面中心的偏差，int16）。阈值放config.py。blob_track.py：find_blobs取最大色块 cx,cy。digit_recog.py：KPU 模型或模板匹配，出class_id+confidence；连续 N 帧一致才采信（防误识，呼应 KB08 FSM）。config.py：所有阈值/ROI/曝光/串口波特(115200)集中放，标"现场标定"。- 复制
contracts/protocol.py到vision/。 - 门：
README写运行步骤 + 镜像版本须与队伍锁定一致；阈值现场标（见 test-checklist）。
与主控对接
- 帧格式由
contracts/protocol.py决定；改协议找 lead 改protocol.yaml重生成，不要在 K230 端硬编。 - 联调：K230 发 → 主控
k230_uart.c的proto_parse_byte收，串口助手可旁观AA 55 .. 0D。
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 · 39 lines · 99 tokens per session scan A 491aa7717bfe
vision-scaffold is a skill published in the GitHub repository Jinchen-Yang/diansai-skill (9 stars, last pushed 1mo ago), licensed MIT. It adds 99 tokens to every session and 782 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 skills, from other repositories
mspm0-ccs
Tool-neutral CLI agent rules for TI MSPM0 development with Code Composer Studio, Keil/uVision, CMake/GCC/OpenOCD, SysConfig, and DriverLib. Use when an agent needs to inspect or modify MSPM0 projects, edit .syscfg configuration, avoid generated SysConfig/build files, use DriverLib APIs, validate SysConfig output…
mspm0-skill
Tool-neutral CLI agent rules for TI MSPM0 development with SysConfig, DriverLib, CCS, Keil/uVision, CMake/GCC/OpenOCD, and supported board references. Use when an agent needs to inspect or modify MSPM0 projects, validate SysConfig output, package examples, or work on NUEDC embedded firmware.
llama-cpp
Runs LLM inference on CPU, Apple Silicon, and consumer GPUs without NVIDIA hardware. Use for edge deployment, M1/M2/M3 Macs, AMD/Intel GPUs, or when CUDA is unavailable. Supports GGUF quantization (1.5-8 bit) for reduced memory and 4-10× speedup vs PyTorch on CPU.
iot-expert
Expert-level IoT systems, embedded devices, edge computing, and IoT protocols. Use when the user mentions embedded, edge computing, MQTT, sensors, or firmware, or when the task involves IoT Architecture, IoT Protocols, Embedded Systems, or Device Design.
embedded-debugging
Debug embedded firmware on real hardware through the jlink-mcp server — crashes, hangs, peripheral misconfiguration, silent devices. Covers loading ELF symbols for source-level backtraces, supplying an SVD for named peripheral fields, and the halt/read/resume discipline. Use whenever debugging a microcontroller over a…
iot-skills
Use when developing with Raspberry Pi Pico (RP2040) for GPIO, I2C, Wi-Fi, MQTT, or sensor integration using MicroPython. Index of 1 skill: KE3036 Keyes Pico learning kit.