Live survey304 models · 14 providers · 23,592 auditsAudited implement · SAFE →What changedHow we verifyllms.txt

Agent skill · Audited from source

oh-story-dsh

skill · zenstory-ai/oh-story-dshMIT licenceAudited 12 Nov 2026 · source 67879d17ebe0

★ 737pushed todayversion 1.0.0
!CAUTIONGrade C · 78/100 · 25 findings · 67879d1
737repo stars · whole zenstory-ai/oh-story-dsh

Is it safe? With care — CAUTION, grade C (78/100). We read this skill's source at commit 67879d17ebe0 and observed filesystem, network and shell access and no hard-coded secrets. The audit raised 25 findings, 0 critical or high. What it does: 从短剧剧本拆出人物/造型、地点/视图、道具/状态和跨场连续性,写成创作者可读的视觉设定。用户说“拆角色/场景/道具”“做资产设定”“判断复用还是新变体”“更新造型/道具状态”,或拿现成剧本直接做视觉资产准备时使用;不写图片提示词,不生成媒体。

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
Cite

Cite this page

ModelsAtlas. “oh-story-dsh (zenstory-ai/oh-story-dsh)”. Retrieved 12 Nov 2026. https://modelsatlas.com/skills/zenstory-ai/oh-story-dsh

Trust audit

Audit 0.4.1 · source read 12 Nov 2026

Trust score78/100!CAUTION C
0 critical0 high7 medium16 low4 info

CAUTION means the audit found things worth knowing before you trust this skill. Grade C is shared by 105 of 18,389 audited skills.

What the source does

  • FilesystemDeclared10 observations
  • NetworkDeclared7 observations
  • ShellDeclared5 observations
  • DependenciesPinnedNothing floats
  • SecretsNone foundNo keys in source

Five audit layers

LayerWhat we checkResultFindings
L0 ProvenanceRepo owner, licence, archive status, activityPASS2
L1 Static sourceSecrets, path traversal, supply chain, hidden codeWARN21
L2 InstructionsPrompt injection, authority framing, hidden directivesPASS2
L3 InventoryThe files and declarations match the sourcePASS0
L4 SandboxRun in isolation and watch behaviourSKIPPED why0

Gates applied: no behavioural pass.

Findings (25)

MEDIUMNetwork egressnet.raw_ip · packages/dsh-plugin/src/workspace-route.ts:639
F-04 Network egressCWE-200 · CWE-319Layer L1
"frame-ancestors 'self' http://127.0.0.1:* http://localhost:*"
Why
No further detail recorded for this rule.
Fix
—
MEDIUMNetwork egressnet.raw_ip · packages/dsh-plugin/src/workspace-route.ts:703
F-04 Network egressCWE-200 · CWE-319Layer L1
const url = new URL(request.url ?? "/", "http://127.0.0.1");
Why
No further detail recorded for this rule.
Fix
—
MEDIUMObfuscation / stealthobf.zero_width · packages/knowledge/oh-story/skills/story-import/scripts/build_outline_brief.py:202
F-07 Obfuscation / stealthCWE-506 · CWE-94Layer L1
"  " + line for line in summary_lines(summary.read_text(encoding="utf-8").lstrip(""))]
Why
No further detail recorded for this rule.
Fix
—
MEDIUMObfuscation / stealthobf.zero_width · packages/knowledge/oh-story/skills/story-import/scripts/build_outline_brief.py:246
F-07 Obfuscation / stealthCWE-506 · CWE-94Layer L1
"\n---\n\n<!-- 资料:outline-reverse-rules.md -->\n\n" + RULES.read_text(encoding="utf-8").lstrip("").strip()]
Why
No further detail recorded for this rule.
Fix
—
MEDIUMObfuscation / stealthobf.zero_width · packages/knowledge/oh-story/skills/story-import/scripts/check-outline-contract.js:121
F-07 Obfuscation / stealthCWE-506 · CWE-94Layer L1
const text = fs.readFileSync(file, 'utf8').replace(/^/, '')
Why
No further detail recorded for this rule.
Fix
—
MEDIUMObfuscation / stealthobf.zero_width · packages/knowledge/oh-story/skills/story-long-analyze/scripts/inspiration_index.py:80
F-07 Obfuscation / stealthCWE-506 · CWE-94Layer L1
return "" + buffer.getvalue()
Why
No further detail recorded for this rule.
Fix
—
MEDIUMObfuscation / stealthobf.zero_width · packages/knowledge/oh-story/skills/story-long-scan/scripts/aggregate-rank.js:171
F-07 Obfuscation / stealthCWE-506 · CWE-94Layer L1
const text = fs.readFileSync(filePath, "utf8").replace(/^/, "");
Why
No further detail recorded for this rule.
Fix
—
LOWInsecure cryptocrypto.weak_hash · packages/knowledge/video-recap/skills/video-recap/scripts/jianying/writer.py:52
F-09 Insecure cryptoCWE-327 · CWE-338Layer L1
digest = hashlib.md5(usedforsecurity=False)
Why
No further detail recorded for this rule.
Fix
—
LOWFilesystem / pathfs.traversal · packages/dsh-plugin/src/role-provider.ts:26
F-05 Filesystem / pathCWE-22 · CWE-59Layer L1
? resolve(current, "../../knowledge/oh-story/roles")
Why
No further detail recorded for this rule.
Fix
—
LOWFilesystem / pathfs.traversal · packages/dsh-plugin/src/skill-provider.ts:306
F-05 Filesystem / pathCWE-22 · CWE-59Layer L1
? resolve(current, "../../knowledge/oh-story/skills")
Why
No further detail recorded for this rule.
Fix
—
LOWFilesystem / pathfs.traversal · packages/dsh-plugin/src/skill-provider.ts:313
F-05 Filesystem / pathCWE-22 · CWE-59Layer L1
? resolve(current, "../../knowledge/drama/skills")
Why
No further detail recorded for this rule.
Fix
—
LOWFilesystem / pathfs.traversal · packages/dsh-plugin/src/skill-provider.ts:320
F-05 Filesystem / pathCWE-22 · CWE-59Layer L1
? resolve(current, "../../knowledge/novel-to-game/skills")
Why
No further detail recorded for this rule.
Fix
—
LOWFilesystem / pathfs.traversal · packages/dsh-plugin/src/skill-provider.ts:404
F-05 Filesystem / pathCWE-22 · CWE-59Layer L1
? resolve(current, "../../knowledge/video-recap/skills")
Why
No further detail recorded for this rule.
Fix
—
LOWNetwork egressnet.raw_ip · README.md:96
F-04 Network egressCWE-200 · CWE-319Layer L1
保持终端运行,浏览器默认自动打开;没有自动打开就复制终端打印的完整 `http://127.0.0.1:3080/?token=...` 链接访问,首次认证需要链接里的 token。关闭终端会停止服务。安装与启动请使用同一个 dsh 版本。本页默认命令固定使用 npm `latest` 的 `0.2.2-rc.1`(需要 Node.js 24.9+)。本插件接受 `>=0.2.0-rc.2 <0.
Why
No further detail recorded for this rule.
Fix
—
LOWNetwork egressnet.raw_ip · README_EN.md:96
F-04 Network egressCWE-200 · CWE-319Layer L1
Keep the terminal running; the browser opens automatically by default. If it does not, copy the full `http://127.0.0.1:3080/?token=...` link printed in the terminal — first-time authentication needs t
Why
No further detail recorded for this rule.
Fix
—
LOWNetwork egressnet.raw_ip · packages/dsh-plugin/README.md:37
F-04 Network egressCWE-200 · CWE-319Layer L1
保持终端运行。若浏览器未自动打开,访问终端打印的完整 `http://127.0.0.1:3080/?token=...` 链接完成首次认证。
Why
No further detail recorded for this rule.
Fix
—
LOWObfuscation / stealthobf.decode_call · packages/knowledge/drama/skills/short-drama-produce/scripts/fixture_adapter.py:22
F-07 Obfuscation / stealthCWE-506 · CWE-94Layer L1
PNG = base64.b64decode(
Why
No further detail recorded for this rule.
Fix
—
LOWObfuscation / stealthobf.decode_call · packages/knowledge/drama/skills/short-drama-produce/scripts/provider_adapters.py:1425
F-07 Obfuscation / stealthCWE-506 · CWE-94Layer L1
content = base64.b64decode(image, validate=True)
Why
No further detail recorded for this rule.
Fix
—
LOWObfuscation / stealthobf.decode_call · packages/knowledge/drama/skills/short-drama-produce/scripts/provider_adapters.py:1461
F-07 Obfuscation / stealthCWE-506 · CWE-94Layer L1
content = bytes.fromhex(audio)
Why
No further detail recorded for this rule.
Fix
—
LOWObfuscation / stealthobf.decode_call · packages/knowledge/drama/skills/short-drama-produce/scripts/provider_adapters.py:1499
F-07 Obfuscation / stealthCWE-506 · CWE-94Layer L1
content = bytes.fromhex(audio)
Why
No further detail recorded for this rule.
Fix
—
LOWObfuscation / stealthobf.decode_call · packages/knowledge/video-recap/skills/video-recap/scripts/voiceover.py:579
F-07 Obfuscation / stealthCWE-506 · CWE-94Layer L1
output_path.write_bytes(base64.b64decode(audio_data))
Why
No further detail recorded for this rule.
Fix
—
LOWPrompt injectionprompt.credential_read · docs/ARCHITECTURE.md:80
F-01 Prompt injectionCWE-94 · CWE-1427Layer L2
Image and video buttons inject an exact prepare-only `/short-drama-produce` request through the current Session's official `ConversationController`. The Session projection remains `awaiting_confirmati
Why
asks the agent to read credentials
Fix
—
LOWPrompt injectionprompt.credential_read · docs/VALIDATION.md:260
F-01 Prompt injectionCWE-94 · CWE-1427Layer L2
| Real provider | `pnpm test:dsh:real` | Paid provider-compatibility observation for the official DeepSeek model, durable Agent completion, Oh Story Role calls, fiction review, short-drama review, rea
Why
asks the agent to read credentials
Fix
—
INFOInventory / provenanceinv.oversize · docs/images/game-studio-jin-ping-mei.png
F-13 Inventory / provenanceCWE-1104Layer L0
docs/images/game-studio-jin-ping-mei.png
Why
1259109 bytes not read
Fix
—
INFOInventory / provenanceinv.oversize · scripts/demo-fixtures/media/shot-ep001-001-keyframe-v2.png
F-13 Inventory / provenanceCWE-1104Layer L0
scripts/demo-fixtures/media/shot-ep001-001-keyframe-v2.png
Why
1633298 bytes not read
Fix
—

Audit 0.4.1 · source sha 67879d17ebe0 · 12 Nov 2026Report an issue or request a re-scan

Provenance

Who published it, and what the verdict is pinned to

  • Publisherzenstory-ai GitHub organizationOwner of the repository this skill is read from
  • Source folderzenstory-ai/oh-story-dshThe exact folder we audited
  • LicenceMITAs the repository states it
  • Commit pinned for this verdict67879d17ebe0A new audit runs when the source changes
  • Repository statusNot archivedPushed today
  • Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

What it tells the agent

Full source on GitHub ↗

Its description: 从短剧剧本拆出人物/造型、地点/视图、道具/状态和跨场连续性,写成创作者可读的视觉设定。用户说“拆角色/场景/道具”“做资产设定”“判断复用还是新变体”“更新造型/道具状态”,或拿现成剧本直接做视觉资产准备时使用;不写图片提示词,不生成媒体。

---
name: short-drama-assets
description: 从短剧剧本拆出人物/造型、地点/视图、道具/状态和跨场连续性,写成创作者可读的视觉设定。用户说“拆角色/场景/道具”“做资产设定”“判断复用还是新变体”“更新造型/道具状态”,或拿现成剧本直接做视觉资产准备时使用;不写图片提示词,不生成媒体。
license: MIT
---

# 短剧资产拆解

从剧本提取可复用、可持续的视觉事实,统一写入 `剧集/<EP>/视觉设定.md`(每个条目使用二级标题 `## <人物|造型|地点|道具> · <中文名称>`);每次出场先判断身份
与状态,剧本语义、镜头走位和生成提示词由各负责技能处理。

## Quick Start

```text
用 $short-drama-assets 从 EP001 剧本拆完人物、造型、地点、道具和本集状态
```

## 入口与边界

现成剧本可以直接拆资产。先区分:

- **身份**:换一个就不再是同一个人、地或物;
- **变体**:身份不变,但服装、伤势、时段、天气、开合或持有状态改变;
- **镜头瞬态**:姿势、视线、左右手、站位和相机角度,由分镜拥有;
- **故事语义**:知识、目标、关系和情绪由写作/开发拥有,只引用可见后果。

## 工作流

1. 读取当前范围剧本与已确定的项目视觉方向;尚未确定时代锚点时,按原著或剧本事实提炼(见 [时代锚点](references/era-anchors.md)),证据不足标为待决定。
2. 逐场提取实体、可见描述、状态变化与含混指代(见 [出现提取](references/occurrence-extraction.md)),对每次出场判定复用、新变体、新身份或未决(见 [身份与变体](references/identity-vs-variant.md));含混指代不猜。
3. 编写最小可识别设定:先写跨场不变的 `识别锚点`,再写 `本集状态`、变化原因与生效范围;不复制剧情摘要,不擅自补造品牌、文字或私密信息。
4. 持续限制后续动作空间的身体—物件—空间关系写进有生效范围的状态,可独立识别的物件单建道具资产,交分镜逐镜落实(见 [道具与状态](references/prop-and-state.md))。
5. 核对造型/伤势、持物/所有权、道具状态、地点时段/天气/光态与跨集结束状态(见 [连续性变化](references/continuity-delta.md))。
6. 对跨镜出现且不一致会被看出的不变可见事实声明 `连续性锁`:锁面用提示词语言写成最小名词短语,写在所属条目下(服装锁写在 `造型` 条目下),分镜未出时镜头范围写 `入画`(见 [跨镜一致性锁](references/continuity-lock.md))。
7. 提示词正文语言不是中文或条目名易与普通词同形时,按 [人物与造型](references/character-and-look.md) 填写 `画面代称`(`AST-13`)。
8. 整集请求一次处理完整集,末尾集中汇报关键身份决定与未决指代。

## 按需知识

默认只读本 SKILL、当前剧本和视觉方向。按需查阅:

- 阶段边界与规则分级:[阶段契约](references/stage-contract.md)
- 定时代、写时代锚点,或避免把场所/设备/服装无依据写旧:[时代锚点](references/era-anchors.md)
- 从剧本提取资产出现证据:[出现提取](references/occurrence-extraction.md)
- 新身份、复用、变体或未决判定:[身份与变体](references/identity-vs-variant.md)
- 人物识别与造型层次:[人物与造型](references/character-and-look.md)
- 地点地理、视图与固定锚点:[地点与视图](references/location-and-view.md)
- 功能道具、所有权、状态变化与可读文字政策:[道具与状态](references/prop-and-state.md)
- 声音参考的身份与授权边界:[声音方向](references/voice-direction.md)
- 跨场、跨镜、跨集状态交接:[连续性变化](references/continuity-delta.md)
- 跨镜服装/道具颜色、材质锁面:[跨镜一致性锁](references/continuity-lock.md)
- 完成前漏项与一致性核对:[资产审查清单](references/asset-review-checklist.md)

## 完成

所有明确出场均已复用、新建、变体化或标为真实未决,且识别锚点、状态变化与连续性锁已写入 `剧集/<EP>/视觉设定.md`。完成后可转 `$short-drama` 做局部或完整机械核对;图片提示词、分镜、审查和生产不自动开始。

Excerpt, first 2,116 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

Install

Pinned to the audited commit 67879d1

Download folder ↗
npx skills add zenstory-ai/novel-to-game -g -y -s '*' \
git clone https://github.com/zenstory-ai/video-recap-skills.git

Commands as the repository documents them: shown, never run by us. Our verdict covers commit 67879d17ebe0 only.

Works in which agents?

What its own text says, not a test

AgentStatusMore
Claude CodeMentioned in its text, not testedSkills for Claude Code · About Claude Code
CodexMentioned in its text, not testedSkills for Codex · About Codex
CursorMentioned in its text, not testedSkills for Cursor · About Cursor
OpenClawMentioned in its text, not testedSkills for OpenClaw · About OpenClaw
“Mentioned” means the skill’s text names the agent. We have not run it there.

Audit history

Every audit, every source change

  1. !CAUTION C 78 / 100

    first audit · source 67879d17ebe0

README badge

For maintainers · shows the latest audit

ModelsAtlas: CAUTION CMODELSATLASCAUTION · C
[![ModelsAtlas: CAUTION C](https://modelsatlas.com/skills/zenstory-ai/oh-story-dsh/badge.svg)](https://modelsatlas.com/skills/zenstory-ai/oh-story-dsh/)

The badge shows our verdict and grade for this skill, and links back to this page. It changes when the audit does.

Add this badge to your README

[![ModelsAtlas: CAUTION C](https://modelsatlas.com/skills/zenstory-ai/oh-story-dsh/badge.svg)](https://modelsatlas.com/skills/zenstory-ai/oh-story-dsh/)

Questions about oh-story-dsh

What does the oh-story-dsh skill do?
DeepSeek Harness 故事创作插件:一站式集成网文连载、短剧漫剧、小说改游戏与视频解说四大工作台 | Community DeepSeek Harness plugin for web novels, short dramas, games & video recaps
Is oh-story-dsh safe to install?
With care. The audit graded it C (78/100) and found 25 things worth knowing before you trust this skill, listed below with the exact line each was found on.
What can oh-story-dsh access on my machine?
The audit observed that it reaches the network, runs shell commands and reads or writes files. Each of those is consistent with what it says it does. Secrets in the source: none found.
Which assistants does oh-story-dsh work with?
Its documentation mentions claude-code, codex, cursor and openclaw. That is what the text claims, not a compatibility test we ran.
How current is this page?
The grade is for one exact copy of the source (67879d17ebe0), read on 2026-11-12. The repository is watched, and a new audit runs when it changes — this is the first audit.

Sources

Spotted something wrong? Report a correction · Methodology

Keep going in your AI

Opens with this page as the source. We add nothing else to the prompt.