skill-description-audit SKILL.md skillA
skill-description-audit from ninjasln-labs/agent-skills is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add ninjasln-labs/agent-skills/skill-description-audit@v1.38.0Pins v1.38.0 in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 17 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- ninjasln-labs/agent-skills
- Path
skill-description-audit/SKILL.md- Versions recorded
- 5
- Latest version
- v1.38.0, committed Oct 5, 2026
- First version recorded
- Sep 22, 2026
- Size
- 44,992 bytes
- Repository stars
- 3
- License
- none declared
Skill Description Audit. User-invoked only: run it when the user asks, never on your own initiative. When asked to audit a skill's description, cross-validate it against that skill's own body and write a graded report beside the skill — the skill is never edited. The audit is a per-criterion sheet: every criterion is answered in turn and graded into a severity band, so a reader sees what was asked, answered, and how sure the auditor was. It covers frontmatter spec compliance, whether every…
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
# 技能描述交叉验证审计（Skill Description Audit） **范围：** 用户级个人技能（正文以安装位置的 `SKILL.md` 为权威；消费侧通过符号链接指向真源目录，遵循技能库「派生只链不拷」约定）。 **正文结构**：**判据集是主干**——题面/答案空间/`band` 挂在[判据集契约](#判据集契约)的渲染表，**每题的判定方法在外置件 `references/criteria-methods.md`、同序同 ID 对齐**，判定标准只写一次（不另立平行体系）。`A` 组先实算、`B` 组逐题作答。 > **逐题讲解为什么外置**：讲解有三百余行，是本技能正文的大头。外置到 > `references/criteria-methods.md` 后正文回到 500 行以内，`LEN-03` 由红转绿。 > **代价是一次跳转**——本技能是 `disable-model-invocation: true`（仅用户显式调用）， > 正文只在显式调用时才加载，**多一跳的实际成本很低**；而若留在正文里， > 每次激活都要为它付 300 余行的上下文。判据退役时讲解仍自动从外置件消失（同源渲染）。 > > **v1.32.0：撤回了「触发面补全」这条待办——它的前提是错的。** 前一版把 > 「三个专节在 description 零触发词」当成待补的 [中] 发现，写进了本注记。 > **本技能是 `disable-model-invocation: true`（仅用户显式调用），本来就不靠自动触发**； > **触发面类判据是审别人用的，拿它审自己属范畴错误**（相关判据已于 2026-10-05 > 退役：见 `references/criteria-governance.md` §8.21）。用户裁定： > **本技能不是常用技能，不需要常驻触发词。** > ⇒ 「触发面补全」这条待办作废，description 按「用户显式调用时说得清做什么」重写（见下）。 > > **实测校准（2026-10-04，外部语料，`scripts/corpus-trigger-rate.py`）**：外部生态里 > user-invoked 技能**占比极低**（件数与占比以该脚本实算输出为准，**不在此处写死**）⇒ 这一档是 > **窄修补，不是系统性改造**。 > 写进判据集时我一度把它讲成常备路径，实测不支持那个说法，已按实算改回。 > **低触发率不是废判据的理由**：§6.1 的 1/3 是**上限**（防噪声），不是下限（防冷门）。 > > **对 `disable-model-invocation: true` 的技能**，「进入面」类维度**不适用**—— …
Read the whole file at its exact version.
How to install
mdr add ninjasln-labs/agent-skills/skill-description-audit@sha256:c45c34ad1b9e5d68The line at the top of this page pins the label v1.38.0, 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_mridsredqvyuhfdi)
1 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 5 versions of this file, between Sep 22, 2026 and Oct 5, 2026, about one every 3 days.
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| v1.38.0 latest | 2026-10-05 | 94956f0 | 44,992 B | A | view · diff |
| v1.14.1 | 2026-10-02 | 6c50dbe | 42,513 B | A | view · diff |
| v1.14.0 | 2026-10-02 | fae63b1 | 42,535 B | A | view · diff |
| v1.12.0 | 2026-09-29 | 0febb7a | 41,492 B | A | view · diff |
| v1.11.0 | 2026-09-22 | 75c75c0 | 36,904 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 (44992 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
ninjasln-labs/agent-skills · 3 stars · license none declared · pushed 2026-10-05 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_mridsredqvyuhfdi GET https://markdownregistry.com/api/v1/resolve?ref=ninjasln-labs/agent-skills/skill-description-audit GET https://markdownregistry.com/api/v1/blob/c45c34ad1b9e5d68ddce9e61b0a400265695538504356848fb6963d99d6e1f80
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.