DeepSeek Harness 插件包:Anima 风格 LoRA 数据集全流程引擎,覆盖抓图下载 → 去重筛选 → 文字检测修补 → 重排编号 → 打标洗标 → 训练配置。自带 anima_status/anima_stage/anima_job/anima_doctor 四个工具、Agent 预设与图形设置页,另有一套可脱离 DSH 独立运行的 15 阶段 Python CLI。
Install verification: sandbox install ok, web smoke failed (L5) · tested on dsh 0.2.1-rc.2 · last checked Install verification details →
What it doesAI
Anima-style LoRA dataset pipeline engine: scrape, dedup, inpaint text, tag and configure training, with DSH tools and a standalone 15-stage CLI.
- Covers image scraping, dedup, text inpainting, tagging and training config
- Ships anima_status/stage/job/doctor tools plus agent preset and settings page
- Standalone 15-stage Python CLI runs without DeepSeek Harness
Capability tags are AI-inferred or rule-extracted from the README (see badge) — vocabulary-limited, not install-verified. AI = inferred; README = literal keyword from the readme.
Installation
dsh plugin --profile web add github:Nana7mi0721/anima-style-lora-workflow-bundleInstall method: GitHub · sandbox install ok, web smoke failed (L5)
Verification & Compatibility
Grouped by plugin version: each card lists the dsh versions actually tested against that artifact version and their install verdicts (L5 wins). Untested combinations are omitted — absence means untested, never assumed compatible. L1–L3 are a plugin-level static archive (independent of the dsh version — the same verdict on every dsh row): they sit in each row's level grid next to L4/L5, marked "static", while their evidence is stored once per plugin inside that row's "checks & evidence" instead of being duplicated per version.
External credentials (plugin-level): none detected
○ The author declares no dsh version requirement — install is not blocked by version.
| plugin \ dsh | dsh 0.2.1-rc.2 current | dsh 0.2.1-alpha.4 | dsh 0.2.1-alpha.2 |
|---|---|---|---|
| 2916f4 | ✗ | ✗ | ✗ |
2916f4github3 failed2026-10-19
https://github.com/Nana7mi0721/anima-style-lora-workflow-bundle
✗dsh 0.2.1-rc.2current2026-10-19L5 · failed
- • Node.js: v24.14.0
- • DSH: 0.2.1-rc.2
- • Plugin artifact: github · https://github.com/Nana7mi0721/anima-style-lora-workflow-bundle (version unstated)
ID: anima-style-lora-workflow-bundle@2916f4@dsh0.2.1-rc.2 · profile: l4-sandbox
Issued: 2026-10-19 · expires: 2026-10-26
View checks & evidence
L1–L3 are plugin-level static conclusions (baseline, identical across dsh versions): evidence is stored once per plugin.
✗dsh 0.2.1-alpha.42026-10-16L5 · failed
- • Node.js: v24.14.0
- • DSH: 0.2.1-alpha.4
- • Plugin artifact: github · https://github.com/Nana7mi0721/anima-style-lora-workflow-bundle (version unstated)
ID: anima-style-lora-workflow-bundle@2916f4@dsh0.2.1-alpha.4 · profile: l4-sandbox
Issued: 2026-10-16 · expires: 2026-10-23
View checks & evidence
L1–L3 are plugin-level static conclusions (baseline, identical across dsh versions): evidence is stored once per plugin.
✗dsh 0.2.1-alpha.22026-10-14L5 · failed
- • Node.js: v24.14.0
- • DSH: 0.2.1-alpha.2
- • Plugin artifact: github · https://github.com/Nana7mi0721/anima-style-lora-workflow-bundle (version unstated)
ID: anima-style-lora-workflow-bundle@2916f4@dsh0.2.1-alpha.2 · profile: l4-sandbox
Issued: 2026-10-14 · expires: 2026-10-21
View checks & evidence
L1–L3 are plugin-level static conclusions (baseline, identical across dsh versions): evidence is stored once per plugin.
Author badges
Embed the official badges in your README to showcase security & install-test status:
`[](https://www.dsh.so/artifact/anima-style-lora-workflow-bundle/)`Security Report
Two automated layers, neither a manual review: a static scan (per-plugin rule checks) and a model-driven dynamic security audit that covers only the subset static-flagged at warn or above and never executes plugin code.
Static low, dynamic medium — stricter wins → overall medium.
v0.1.0dsh manifest0.1.0git2026-10-210.1.0same versionclean · gitVet is an AST static scan (npm artifact or git source).
show static findings (4)
- warningchildProcess · code-exec
lib/run.js:414const child = spawn(python, argv, { cwd: config.pythonDir, env, windowsHide: true }) - warningchildProcess · code-exec
lib/settings.js:161child = spawn(python, ['-X', 'utf8', '-m', 'animasl.cli', '--home', config.home, ...argv], { - warningnetIp · network
python/animasl/config.py:115"proxy_candidates": ["", "http://127.0.0.1:7897", "http://127.0.0.1:7890", - warningnetIp · network
python/animasl/config.py:116"http://127.0.0.1:10809", "http://127.0.0.1:1080"],
show 3 audit findings
- mediumPrototype pollution in settings write: dotted keys reach Object.prototype via setPath
lib/settings.js:69READ_ONLY_KEYS only blocks the top-level key 'home'. A POST body like {"set":{"__proto__.x":1}} flows through applyPatch -> setPath: for part '__proto__' the guard sees an object and does not replace it, so `node` becomes Object.prototype and the final assignment (or the intermediate one) writes onto the prototype of the whole DSH host process. Any request passing the trust fence (a no-Origin local process, or a same-hostname page using a simple text/plain POST, which needs no CORS preflight) can do this. Effect ranges from host-process denial of service to gadget-assisted code execution depen
- mediumSettings API Origin check ignores port, so any same-hostname page can read stored booru credentials
lib/settings.js:249The fence accepts a request whenever the Origin hostname equals the Host hostname, regardless of scheme or port, and accepts requests with no Origin at all. When the DSH web server is bound to 'localhost' (isLoopbackHostname returns true for it), a page served from http://localhost:<any other port> is 'same-site' (Sec-Fetch-Site is not 'cross-site') and passes the Origin hostname comparison, so it can POST to /anima-lora/api/read and receive the runtime layer of animasl.config.json containing danbooru_login/danbooru_api_key and the three exhentai cookies, and can POST text/plain JSON to /write
- lowprobe method exposes existence of arbitrary filesystem paths without an allowlist
lib/settings.js:348Any caller that passes the trust fence can enumerate up to 64 arbitrary absolute paths per request and learn which files/directories exist on the host. Combined with the fence weakness above, a same-hostname local page could map the user's filesystem. No file contents are returned and a local process already has equivalent access, so impact is limited to information disclosure.
DSH bundle whose local settings HTTP API can be reached from any same-hostname origin and whose config writer is vulnerable to prototype pollution; no remote code execution or shell injection found.
Rule: the dynamic audit can only raise risk, never lower it (raise-only, never an exemption); a wrong raise is withdrawn by human review and the tier falls back to the static verdict. Overall verdict = the stricter of the two layers.
Disclaimer: the static scan and the dynamic security audit are both automated analysis, not a security guarantee. Always review what you install.
Activity
Last commit 2026-10-20 · activity: Active
• Repo created: 2026-10-04
• Stars: ★ 0 · Forks: 0
• Health: Active — committed within last 30 days
Source
GitHub: github.com/Nana7mi0721/anima-style-lora-workflow-bundle
Author README
View on GitHub ↗The author's original README from the repository — fetched live, unedited (the English page loads the default README).
