ai-cross SKILL.md skillA
ai-cross from keros68/xiaoyu-skill is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add keros68/xiaoyu-skill/ai-cross@v1.21.3Pins v1.21.3 in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 17 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- keros68/xiaoyu-skill
- Path
skills/ai-cross/SKILL.md- Versions recorded
- 12
- Latest version
- v1.21.3, committed Oct 7, 2026
- First version recorded
- Oct 1, 2026
- Size
- 17,342 bytes
- Repository stars
- 14
- License
- MIT
多模型分工与跨厂商交叉验证 skill：把任务派给合适的模型分工执行，用不同厂商的模型互相核查关键产出，按档位分层派发以节省订阅额度。适用场景：用户要求派发任务、分层执行、多模型协作、对关键产出做交叉验证、盘点或接入可用模型，或提到 dispatch、派工；也包括日常说法：让别的模型或另一家看看、换个模型审一下、要个第二意见、找别家复查、多找几家对一下。不适用场景：没提到其他模型的普通任务（只说"帮我检查一下"不算，没有派发需求时不要触发）、没有 shell 执行能力的纯聊天宿主。.
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
# ai-cross — 多模型分工与跨厂商交叉验证 **核心价值：跨厂商交叉验证——用不同厂商的模型互相挑错。** 同厂商多 agent 的复查能抓随机错误与上下文污染，但失败模式相关；抓宿主模型自己的系统性盲区，需要另一家厂商的独立视角。其余能力是围绕它的支架：分层派发（粗活给便宜档，贵额度留给关键判断）、熔断、留痕、闭环。 **第 0 步，永远先做：查环境变量 `AI_CROSS_PEER`。为 1 → 你自己就是被派出的子任务：只执行任务、直接作答，跳过本 skill 的全部派发流程**（防套娃；这条不放最前面就会被漏掉）。 本文件是入口和每次都要做的判断。完整规则在 `references/rules.md`，按文末「什么时候读哪一节」按需读，不必预先通读。 ## 从这里开始 1. 先看这台机器盘点过没有：`python <本skill目录>/references/inventory.py show --human`。manifest 存在用户主目录下（`$AICROSS_HOME/skill/manifest.json`，未设时 `~/.aicross/skill/manifest.json`），各宿主共用一份，skill 更新不影响它——**每台机器盘点一次，换宿主不用重做**。 2. 退出码 3（还没有 manifest）→ **直接进入盘点**：读 `references/setup.md` 按其执行。**被点名且没有 manifest 时，盘点就是当前任务——不要只复述规则等用户开口。** - **但盘点只是加分项，不是执行的前置**：用户明说"先别盘点，就把这个干了"时，退化到宿主内部通道（Claude Code 有 scout/worker/heavy）或用户当场指定的通道执行，并在路由决策行注明「未盘点 → 跨厂商验证不可用」。**缺盘点只降级能力，不阻塞任务。** - **用户是带着具体任务来的**（"让别的模型看看这段代码"），不是来盘点的：先说一句"这台机器还没盘点过，先花一分钟登记本机的模型"，盘点做到保存为止（setup.md 第 1–2 步），然后回到用户的任务——不单独冒烟，这次派发就是被选中那一条的第一次实测，成功后脚本自动记为通过；任务收尾时补上 setup.md 第 5 步的报告：清单存在哪，并把「以后怎么用」那张表原样给用户。 - 没有 Python 跑不了脚本时，读 `~/.aicross/skill/manifest.md`（手工清单，写法见 setup.md「兜底」）。 3. 已有 manifest → 输出里就是「厂商 × 档位」清单、默认审查顺序和默认执行者。模型 ID 只用清单里的，不凭记忆写。 …
Read the whole file at its exact version.
How to install
mdr add keros68/xiaoyu-skill/ai-cross@sha256:fd2fffbcef2770c8The line at the top of this page pins the label v1.21.3, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
[](https://markdownregistry.com/a/art_l5witbz6amftfrcz)
3 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 12 versions of this file, between Oct 1, 2026 and Oct 7, 2026, more than one a day.
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| v1.21.3 latest | 2026-10-07 | 9001718 | 17,342 B | A | view · diff |
| v1.21.2 | 2026-10-07 | 3257884 | 17,280 B | A | view · diff |
| v1.21.1 | 2026-10-05 | 45f3a41 | 17,009 B | A | view · diff |
| v1.21.0 | 2026-10-05 | dcbbe7d | 16,731 B | A | view · diff |
| v1.20.0 | 2026-10-05 | 217cb8c | 55,609 B | A | view · diff |
| v1.19.1 | 2026-10-05 | 6ca4f84 | 50,606 B | A | view · diff |
| v1.19.0 | 2026-10-05 | 36793dd | 48,945 B | A | view · diff |
| v1.18.2 | 2026-10-03 | 7791cd7 | 47,063 B | A | view · diff |
| v1.18.1 | 2026-10-03 | d19c13c | 47,063 B | A | view · diff |
| v1.18.0 | 2026-10-01 | 622c6f8 | 47,063 B | A | view · diff |
| v1.17.2 | 2026-10-01 | f305eb1 | 47,063 B | A | view · diff |
| v1.17.1 | 2026-10-01 | 236ff9b | 47,063 B | A | view |
Audit of the latest version
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (17342 bytes)
- pass: No zero-width or bidi control characters
- pass: No instruction hidden inside an HTML comment
- pass: No link to an exfiltration or paste host
- pass: No credential-shaped string
- pass: No instruction to send local credentials anywhere
- pass: No text hidden with inline styles
- pass: No prompt-injection phrasing
- pass: No curl or wget piped into a shell
- pass: No recursive delete of root, home or parent
- pass: No instruction to read or print local credentials
- pass: No base64 blob over 200 characters
- pass: No link to a raw IP address
- pass: No script tag
Source
keros68/xiaoyu-skill · 14 stars · license MIT · pushed 2026-10-22 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_l5witbz6amftfrcz GET https://markdownregistry.com/api/v1/resolve?ref=keros68/xiaoyu-skill/ai-cross GET https://markdownregistry.com/api/v1/blob/fd2fffbcef2770c89b31b2c1dbdd7452a00357c41d40c56c4cadb1f8e447aa44
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.