world-model-runtime SKILL.md skillA
world-model-runtime from ooooooooooooooooooop/personal-ai is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add ooooooooooooooooooop/personal-ai/world-model-runtime@git:20261022.b7059fePins git:20261022.b7059fe in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 17 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- ooooooooooooooooooop/personal-ai
- Path
skills/world-model-runtime/SKILL.md- Versions recorded
- 8
- Latest version
- git:20261022.b7059fe, committed Oct 22, 2026
- First version recorded
- Sep 13, 2026
- Size
- 11,022 bytes
- Repository stars
- 4
- License
- MIT
在需外置持久认知状态的任务上运行世界模型闭环的执行协议——恢复持久状态→建立竞争模型→显式预测→真实观察→修订模型→主动取证→决策→持久化。仅用于跨会话继续、恢复开放预测、持续竞争模型、需向 canonical 提交治理提案的场景；常规代码编写/Bug调试/一次性任务严格保持 OFF（反触发词：跨组件≠CORE；mutation≠CORE；debug难≠CORE）；加载后按 Gate V0.2 判据自判形式化深度。.
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
# 世界模型运行时 你正在执行一个带持久世界模型的任务。世界模型不是记忆库，是「当前对世界如何运作的最佳压缩 + 主动被测试的对象」。本协议定义九环如何真实执行，不是形式主义清单。 ## 模式（Gate V0.2 权威判据） **加载条件**（本 skill 何时该被用）与**形式化深度**（OFF/CORE/FULL）是两件事，不要混： - **加载条件**：仅用于决策关键状态无法被当前会话或普通工件（代码/Git/测试输出）承载、必须外置持久认知状态的场景；常规代码开发与排查不加载。 - **形式化深度**：依据 Gate V0.2 判据（`externalized-state demand` 轴）**默认自判为 OFF**。 **判据原则（初测观察提示盲开形式化门禁会导致严重的执行延迟与交互摩擦，待冷启动对撞实验量化复现）**： - **默认原则**：默认严格保持 **OFF**；只有存在强外置状态需求才升 **CORE**；只有需主动认知控制才升 **FULL**。 - **反触发规则（Anti-Triggers）**： > `跨组件 ≠ CORE；mutation ≠ CORE；debug 难 ≠ CORE` > 普通代码修改、复杂 Bug 排查、跨多个文件的调用、测试运行，状态完全由代码库和会话上下文承载，**严禁升入 CORE**。 - **OFF（默认基线）**：当前会话足以容纳关键状态；任务一次性自包含；状态丢失重建成本低。常规日常开发一律保持 OFF，不触发任何认知门禁。 - **CORE（外置状态需求）**：必须同时满足以下条件之一： 1. 跨阶段/跨会话继续，且必须恢复上一会话未决的 `open prediction`； 2. 必须在多轮间持续维护显式的竞争因果模型（Competing Models H1/H2）； 3. 涉及可能导致灾难性后果的系统级不可逆变更（需前置绑定预测）； 4. 产出必须向 canonical 提交治理提案（proposals）的长期结论。 - **FULL（主动认知控制）**：在 CORE 基础上，还需主动进行信息增益最高的区分性探针（probe）、H4 假设与本体结构修订、长期价值权衡与 Goodhart 检查。 FULL 中每一环标 `USED / NOT_NEEDED(reason) / UNKNOWN`，允许空环但不允许跳过评估。 > 注：形式化深度**不是开关**——世界模型永远在场，OFF/CORE/FULL 只是仪式深度。 > 另有插件侧的静态 `config.mode`（`dsh/world-model/cordis.patch.yml`）， > 那是**守卫严格度**的部署配置，与本协议的会话内自判不是同一件事。 ## 不变量 …
Read the whole file at its exact version.
How to install
mdr add ooooooooooooooooooop/personal-ai/world-model-runtime@sha256:47e8c3270d6a590aThe line at the top of this page pins the label git:20261022.b7059fe, 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_rbda3pmkjqjsfv4h)
3 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 8 versions of this file, between Sep 13, 2026 and Oct 22, 2026, about one every 6 days.
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| git:20261022.b7059fe latest | 2026-10-22 | b7059fe | 11,022 B | A | view · diff |
| git:20261010.bd7b582 | 2026-10-10 | bd7b582 | 9,988 B | A | view · diff |
| git:20261010.1e898e2 | 2026-10-10 | 1e898e2 | 9,466 B | A | view · diff |
| git:20261010.e719f37 | 2026-10-10 | e719f37 | 9,218 B | A | view · diff |
| git:20261010.3739836 | 2026-10-10 | 3739836 | 9,004 B | A | view · diff |
| git:20260918.ef73537 | 2026-09-18 | ef73537 | 7,449 B | A | view · diff |
| git:20260914.0c9411e | 2026-09-14 | 0c9411e | 7,387 B | A | view · diff |
| git:20260913.97c758e | 2026-09-13 | 97c758e | 7,405 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 (11022 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
ooooooooooooooooooop/personal-ai · 4 stars · license MIT · pushed 2026-10-23 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_rbda3pmkjqjsfv4h GET https://markdownregistry.com/api/v1/resolve?ref=ooooooooooooooooooop/personal-ai/world-model-runtime GET https://markdownregistry.com/api/v1/blob/47e8c3270d6a590a0f455d8520ccd55673f1b769ddef559ad1ed3e5d96ff5d1e
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.
More from ooooooooooooooooooop/personal-ai
Every file in ooooooooooooooooooop/personal-ai