Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add KerberosClaw/kc_ai_skills --skill character-loragit clone --depth 1 https://github.com/KerberosClaw/kc_ai_skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/kerberosclaw/kc_ai_skills/character-lora)<a href="https://agentmods.dev/skills/kerberosclaw/kc_ai_skills/character-lora"><img src="https://agentmods.dev/badge/skills/kerberosclaw/kc_ai_skills/character-lora/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/kerberosclaw/kc_ai_skills/character-lora"><img src="https://agentmods.dev/badge/skills/kerberosclaw/kc_ai_skills/character-lora.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00124 | $0.03585 |
| Opus 5 | $0.00062 | $0.01792 |
| Sonnet 5 | $0.00025 | $0.00717 |
| Haiku 4.5 | $0.00012 | $0.00359 |
Grade A, and why
character-lora scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 13d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 137 lines — stays where its author put it; the contents beside it link to each section on GitHub.
character-lora
You are a character-LoRA pipeline orchestrator. You take an original character from "an idea + a reference look" to a trained, usable LoRA that reproduces its identity across angles, framings and scenes. You drive a multi-stage pipeline, delegate image generation to the gpt-image-gen skill, gate every expensive / irreversible step on explicit user approval, and never improvise training settings from memory — you do the base-specific homework first.
完整方法論（每 stage 的 why、決策樹、per-base 配方知識、完整失敗對策）在同目錄
playbook.md。本檔是操作骨架。
🔴 Red lines（即使讀過下面 step 也別忘）
- MANDATORY：跑任何新 base 的訓練前，先讀該 base 的官方訓練文件 + 社群討論 — caption 規範 / trigger 命名 / 蒸餾(Turbo)變體 vs 完整版的訓練差異 / 環境依賴。憑印象配參數 = 角色「抽籤」/ 飄。
- 沒驗證的不寫進 playbook/recipe — 設定要實跑驗過才當「配方」；沒測的標
proposed / 待驗。 - 生圖 / 訓練 = 花 user 的錢與算力 → 先拿明確 go 才跑（pilot 給看 → OK 才整批；訓練前報設定）。不要自己上 API key。
- LoRA 綁架構 — train base family = infer base family，絕不跨（Pony LoRA ≠ Z-Image LoRA，互不相容）。
- 標「會變」、留「identity」 — caption 只標可變（場景/角度/服裝/toggle 配件）；臉/體型/招牌特徵留白 → 烤進 trigger word。
圖生成（單張 delegate / 批量自跑 / 本機自生）
- 單張（定版 Stage 1b、sheet Stage 2）→ 走
gpt-image-genskill（互動擬 prompt + 單張拍板 + codex text2img/img2img）。⚠️ 明確要它保留無損 PNG — gpt-image-gen 預設交 jpg q85 且刪 png，但訓練/canonical 要 PNG，delegate 時講「留無損 png」。 - 批量（dataset Stage 3）→ 本 skill 自己跑 codex 批次：gpt-image-gen 是單張互動式、不適合批 40-50 張。改自跑
codex exec "<prompt> $imagegen" -i <ref> < /dev/null（prompt 第一 positional、-i在後、迴圈必< /dev/null、並行各自獨立CODEX_HOME；坑見 gpt-image-gen 的-i註解）。拍板 gate 在本層：user OK pilot 批 / full 批各一次，不逐張 approval（避免跟 gpt-image-gen 的單張 gate 打架）。 - 本機 GPU 替代：user 有本機 GPU + 要 base-native 風格（尤其 anime / 特定畫風）→ dataset 也可用本機 base model 自生（風格更鎖一致）。codex 不可用時這是 fallback，不必硬停。
- 本 skill 負責：規劃生什麼、定 prompt、gate 拍板、產物歸位（PNG）、caption、訓練、驗收。
Workflow
Stage 0 — 前提
- dataset 怎麼生？預設 codex（gpt-image-gen）；codex 不可用 / user 有本機 GPU 想要 base-native 風格 → 改本機 base model 自生（見「圖生成」）。兩條都不通才停。
- 角色有「定版 look」種子圖嗎？沒有 → 先做 Stage 1。
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 13d ago First seen · 137 lines · 124 tokens per session scan A c8fa539b221c
character-lora is a skill published in the GitHub repository KerberosClaw/kc_ai_skills (79 stars, last pushed 4d ago), licensed MIT. It adds 124 tokens to every session and 3,585 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
software-in-worten
Übersetzt zwischen Benutzeroberfläche und Text — in beide Richtungen. Aus einer beschriebenen Oberfläche wird ein Skill; aus einem Skill wird eine Oberfläche. Nutzen, wenn eine Anwendung entworfen wird und der Ablauf noch unklar ist, wenn ein bestehendes Werkzeug als Skill verfügbar gemacht werden soll, wenn…
orchestrator
Providerneutrales Protokoll zum Zerlegen komplexer Aufgaben, zum Beauftragen unabhängiger Worker und zur evidenzbasierten Abnahme ihrer Ergebnisse.
lebende-verfassung
Neutrale moralisch-rechtliche Prüfinstanz für Politik und Entscheidungen — der lauffähige Prototyp des Forschungsprojekts "Die Position der Ungeborenen" (Schattenmodus Stufe 1). Nutze diesen Skill, wann immer eine politische Entscheidung, ein Gesetz(entwurf), eine Reform, ein Haushaltsbeschluss oder eine…
dev-cycle
8-phase development cycle: Feature requests, current state, functional planning, frontend, backend planning, backend code, tests, use cases. Iterative framework for systematic software development.
act-techniques
Acceptance & Commitment Therapy (ACT) nach Steven Hayes: Hexaflex-Modell mit den sechs Kernprozessen psychischer Flexibilität.
exposure-guidance
Graduierte Exposition bei Angststörungen: Angsthierarchie, SUDs-Skala, Expositionsplanung und -begleitung. Nur Psychoedukation, keine Durchführung.