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 KhazixW2/Everything-Maa --skill maa-pipeline-guidegit clone --depth 1 https://github.com/KhazixW2/Everything-MaaWrote 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/khazixw2/everything-maa/maa-pipeline-guide)<a href="https://agentmods.dev/skills/khazixw2/everything-maa/maa-pipeline-guide"><img src="https://agentmods.dev/badge/skills/khazixw2/everything-maa/maa-pipeline-guide/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/khazixw2/everything-maa/maa-pipeline-guide"><img src="https://agentmods.dev/badge/skills/khazixw2/everything-maa/maa-pipeline-guide.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00080 | $0.05828 |
| Opus 5 | $0.00040 | $0.02914 |
| Sonnet 5 | $0.00016 | $0.01166 |
| Haiku 4.5 | $0.00008 | $0.00583 |
Grade A, and why
maa-pipeline-guide 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 10d 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 — 468 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Universal Pipeline 编写指南
官方知识核对
本指南中的 Pipeline 约定是工作经验和社区规范，不能替代 pinned 版本的官方协议、schema 或源码。当字段语义、默认值、版本差异或 API 行为存在疑问时，通过 $maa-wiki 定位 MaaLLMWiki catalog 中的原始来源，再以官方文档、tools/pipeline.schema.json 或 MaaFramework 源码为准。
如果用户提出的是尚未定义起始状态、安全边界和验收条件的端到端自动化目标，先交给 $maa-workflow-build 建立任务契约与状态机；已有契约时，再用本 skill 处理 Pipeline 设计、修改或审查。
项目初始化接力
开始广泛扫描仓库前，先在目标项目根目录查找 basic_info.md：
- 存在时先读其中“0. Maa Skills 接力协议”，再优先读第 3/4/5/7/8/9 节，获取主 Pipeline、公共节点、返回/弹窗、OCR、模板与 ROI 约定。
- 它只是
maa-project-init生成的上下文缓存；待修改节点必须回到当前 JSON/Python 核实，设备相关结论必须用当前截图或识别结果核实。 - 文件缺失或没有第 0 节时，按本 skill 正常发现项目结构并说明未使用初始化缓存；不得自动调用
$maa-project-init，只有用户明确要求初始化或刷新时才调用。 - 相关
interface.json、Pipeline 或 Agent 文件晚于basic_info.md时，将缓存视为可能过期并以当前源码为准；不得自动刷新或覆盖已有非空文件。
核心原则
- 状态驱动：遵循"识别 → 操作 → 识别"循环。每次操作必须基于识别结果，禁止假设操作后画面状态。
- 高命中率：扩充
next列表，覆盖当前操作后所有可能画面，力争一次截图命中。 - 显式等待策略：优先通过中间识别节点确认状态，不用盲目的长
delay掩盖问题；但启动、动画、结算、加载稳定等场景可以使用短的pre_delay/post_delay/timeout/*_wait_freezes。当确实不需要等待时，要在节点上显式将rate_limit/pre_delay/post_delay设为 0（协议默认rate_limit=1000ms、pre_delay/post_delay=200ms，省略字段会引入隐式等待）。不要假设仓库存在自动补默认值脚本，使用前先发现真实工具。 - 720p 基准：所有坐标、ROI、图片必须基于 720X1280。
- 点击目标由识别结果推导：优先让 MaaFramework 点击识别框中心（
target默认true）。不要用DirectHit+ 硬编码target跳过识别，也不要在识别命中后用写死的target_offset把点击挪到未被识别的控件上——模拟器分辨率、缩放与 UI 版本差异会让这类坐标静默点错位置。 - 格式化：JSON 遵循
.prettierrc（4 空格缩进，数组元素换行）。
需要完整字段速查时读取 references/field-reference.md，不要把整份字段表重复加载到日常任务上下文。编写或审查带点击动作的节点时读取 references/coordinate-hygiene.md，其中有两种坐标坏味、OCR/TemplateMatch 点击的推荐写法和允许偏移的例外条件。
历史审查后的设计准则
这些规则来自 MaaGumballs 与 M9A 的 Pipeline 历史审查，优先级高于早期经验里的绝对化表述：
- 状态机优先，不等于禁止 Python：稳定、可枚举的页面流转优先写成
next+[JumpBack]。当逻辑需要运行时数据、事件库、动态目标选择、跨节点计数、复杂 OCR/图像后处理、pipeline_override计算或失败策略时，使用 CustomAction/CustomRecognition。 - Custom 不只是 action：MaaGumballs 主要使用
action: Custom，M9A 同时大量使用custom_action、custom_recognition、tasker_sink。设计新流程时先判断问题属于“控制流/动作决策”还是“识别/列表解析/图像后处理”。 - 链路要显式：父节点的
next放“当前页面可能出现的下一批状态”；临时弹窗、加载、确认框用[JumpBack]；高风险分支（战斗、购买、消耗、结算继续）要和普通调查/领取/返回分开。 - 等待不是禁用项：不要用盲目的长
delay掩盖状态识别问题；但启动、切页动画、结算、加载后稳定画面等场景可以使用短的post_delay、rate_limit或*_wait_freezes，并配套下一屏识别验证。 - 校验分层：资源加载通过只说明 JSON/资源可加载，不代表 Custom 名称、Python 参数路径、
run_task()结果判断都正确。Custom 映射和关键链路需要单独检查。
What ships with it
3 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.
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.
- 10d ago First seen · 468 lines · 80 tokens per session scan A d8188c63d203
maa-pipeline-guide is a skill published in the GitHub repository KhazixW2/Everything-Maa (12 stars, last pushed 2d ago), licensed MIT. It adds 80 tokens to every session and 5,828 once invoked, about $0.0004 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 2026-08-30.
Other skills, from other repositories
playwright-cli
Automate browser interactions, test web pages and work with Playwright tests.
hive.slack-notifications-setup
Set up a Slack notification channel (Sentinel) for a colony by driving the browser — reuse or create the "Hive Sentinel" Slack app from a JSON manifest, install it, capture the bot + app tokens, create/select the channel via the Slack API, and turn Sentinel on so the colony can ping the user on Slack and accept…
amazon-search-listing
Amazon search and category listing scraper: extract product listings from any Amazon search results page, keyword search URL, or category browse page and return per-item cards (asin, title, url, image, price, listPrice, stars, reviewCount, badges, isAmazonChoice, isBestSeller, isSponsored, delivery, boughtInPast…
etsy-product-detail
Etsy product detail scraper: given an Etsy listing URL, returns full product detail including listingId, title, priceCurrent, priceOriginal, currency, images (all), description, shopName, shopUrl, rating, reviewCount, favorites, inCartCount, variations (with per-option price ranges), highlights, listedDate…
walmart-product-detail
Walmart product detail page extractor: given a walmart.com product URL (walmart.com/ip/...), extract full product data including itemId, title, brand, model, UPC, price, wasPrice, currency, availability, category path, seller info, all images, shortDescription, longDescription, product highlights, full specifications…
amazon-bestseller-listing
Amazon Best Sellers listing scraper: extract product cards from any Amazon Best Sellers (zgbs) or /gp/bestsellers/ category page — returns rank (position on chart), asin, title, url, image, imageAlt, price, stars, reviewCount, ratingRaw per item, plus category metadata (categoryName, categoryFullName, categoryUrl) and…