Not in the top 1,000 by GitHub stars
Xerxes-2/tavern2agent
Compile SillyTavern character cards into pi-native, event-driven interactive narrative runtimes.
History
Review weekly changes to SKILL.md before updating. Dates mark the weeks in which we observed each version; checks apply to the recorded commits.
View weekly changes
SKILL.md
Week of Nov 2, 2026 · +0 / −0 lines
Before Network or broad shell → After Network or broad shell
601b1c3f447a → 601b1c3f447a
Week of Oct 26, 2026 · Line counts unavailable
Before Not checked yet → After Network or broad shell
First observation → 601b1c3f447a
No earlier version was observed for comparison.
Heat
0.4- Star growth, last 7 daysNot available
- Star growth, last 24 hoursNot available
- Rank movement, last 7 daysNot available
- Commits, last 30 daysNot available
- Accounts asking about it through MCP, last 7 daysNot available
- Hot votes this weekNot available
Heat never changes the order by stars, and sponsorship never touches it. How heat works
Stars over time
30-day change not yet available: not enough history.
Individual skills in this repository
How it works
tavern2agent
When your agent uses it
用户提供 SillyTavern 角色卡(PNG/JSON)并要求转换、迁移、移植到 pi coding agent 时使用;覆盖纯 prompt、世界书、MVU、骰子、战斗、好感度、经济、隐藏信息、多 agent 场景。
What it loads into context
Frontmatter
85 tokensAlways available to the agent
SKILL.md
3,216 tokensWhen the skill is activated
Reference documents
46,563 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 开工
- 探索命令
- 信息源 → IR
- 方案
- 多 agent 判定
- Reference 路由
- 产出
- 硬约束
- 完工
Scripts executed separately
scripts/extract_card.pyscripts/get_entry.pyscripts/list_entries.pyscripts/start.sh
Files
.gitignoreasset29 bytesLICENSEasset1,065 bytesREADME.mddoc5,689 bytesREADME_ZH.mddoc4,893 bytesSKILL.mdcore11,825 bytesSKILL.mdcore11,825 bytesSKILL.mdcore11,825 bytesdocs/developing-cards.mddoc2,639 bytesdocs/epistemic-audit-2026-07.mddoc12,047 bytesdocs/tooling.mddoc1,713 bytesreferences/card-ir.mddoc4,034 bytesreferences/codeact.mddoc3,280 bytesreferences/data-layer.mddoc4,517 bytesreferences/decision-tree.mddoc4,799 bytesreferences/design-principles.mddoc4,241 bytesreferences/engineering-discipline.mddoc6,556 bytesreferences/event-packs.mddoc5,141 bytesreferences/evented-runtime.mddoc8,481 bytesreferences/multi-agent-architecture.mddoc11,682 bytesreferences/mvu-mapping.mddoc9,569 bytesreferences/pi-integration.mddoc7,297 bytesreferences/prompt-composition.mddoc10,119 bytesreferences/script-analysis.mddoc2,499 bytesreferences/setup.mddoc2,534 bytesreferences/state-schema-migrations.mddoc2,771 bytesreferences/storytelling.mddoc1,314 bytesreferences/tool-abstraction.mddoc21,580 bytesreferences/ts-engine.mddoc8,816 bytesreferences/two-pass-rendering.mddoc11,692 bytesreferences/validation.mddoc13,485 bytesscripts/extract_card.pyruns9,752 bytesscripts/get_entry.pyruns2,087 bytesscripts/list_entries.pyruns3,242 bytesscripts/start.shruns3,784 bytes
Read SKILL.md
# Tavern → Agent
把 SillyTavern 卡编译成 pi-native 互动叙事 runtime。还原作者想做的游戏:prompt 描述世界,领域事件改变世界,engine/reducer 维护正确性,session 保存存档。不复刻 ST 宏、COT、JSON Patch、HTML 状态栏,不把状态栏字段搬进 `patch_state`。
## 开工
1. 解包卡片,确认输出目录。目录存在时先问:覆盖、增量、另建?
2. 先读 `references/evented-runtime.md`、`references/card-ir.md`、`references/event-packs.md`、`references/design-principles.md`。
3. 全量审计 `data` 字段、世界书、TH scripts、regex scripts、开场白。
4. 先输出或草拟 `world-data/card-ir.json`;不要直接从原卡文本生成代码。
5. 从 IR 形成 Runtime Plan:archetype、event packs、state roots、visibility policy、fact sources、tool surface、subagent roles、prompt modules、validation plan。
6. 写代码前先给用户看 Runtime Plan;复杂卡还要给 state schema、event catalog、reducer/API 清单、事实源边界和子代理边界。
增量更新:先看 `git log -20` + `git diff`。只改本次需求相关文件;不碰 `sessions/`、`runtime/`、`.pi/agent/`。View the complete source on GitHubSkills included
- tavern2agent
Repository facts
- License
- MIT
- Created
- May 10, 2026
- Last pushed
- Jul 11, 2026
- Forks
- 8
- Open issues
- 0
- Project activity
- Slowing
Before you install
Network or broad shellAutomated signals, not a guarantee
Checked Nov 6, 2026
Not fully analyzed: some files were too long, unreadable or missing. This never lowers the level.
Show 2 notes
SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzed
- 1 broad permissionin 1 file: SKILL.md:1
- No dangerous shell commands
- 12 network referencesin 5 files: README.md:34 README_ZH.md:34 scripts/start.sh:6 and 2 more
- 2 obfuscation signalsin 1 file: scripts/extract_card.py:82
- No instruction overrides
- 4 executable scriptsin 4 files: scripts/extract_card.py:1 scripts/get_entry.py:1 scripts/list_entries.py:1 and 1 more
- No access to secrets
- No requests to vote in Hot
Show all 19 signals by file
README.md
- Network accessline 34
Using [`npx skills`](https://github.com/vercel-labs/skills) (recommended): - Network accessline 43
git clone --depth 1 https://github.com/Xerxes-2/tavern2agent \ - Network accessline 126
This skill's methodology is continuously validated and refined in [fate-sandbox](https://github.com/lolo-s-Cosmos/fate-sandbox), a Type-Moon universe sandbox spanning 13 timelines, including FSN Fuyuki, strange Fake, Tsukihime, and The Gard
README_ZH.md
- Network accessline 34
用 [`npx skills`](https://github.com/vercel-labs/skills)(推荐): - Network accessline 43
git clone --depth 1 https://github.com/Xerxes-2/tavern2agent \ - Network accessline 126
本技能的方法论在 [fate-sandbox](https://github.com/lolo-s-Cosmos/fate-sandbox) 上持续验证迭代——一个型月世界观沙盒,13 条世界线(FSN 冬木、strange Fake、月姬、空之境界等)。这里的两段式结算/渲染拆分、引擎台账、确定性 compaction、缓存友好渲染历史等经验均出自它的长跑实测。想看一个完整的 evented standard + 多 subagent + 两段式项目长什么样,读它的源码和
SKILL.md
- Broad permissionsline 1
Bash, Read, Write, Edit, Glob, Grep
references/mvu-mapping.md
- Network accessline 31
- 框架 API / 事件 / `stat_data` 存储 / `parseMessage`:`https://github.com/StageDog/tavern_helper_template/raw/refs/heads/main/.cursor/rules/mvu变量框架.mdc` - Network accessline 32
- schema 契约(zod 4、幂等增量解析、transform 软约束)/ 世界书条目结构(`[initvar]`、`变量列表`、`[mvu_update]变量更新规则/变量输出格式`):`https://github.com/StageDog/tavern_helper_template/raw/refs/heads/main/.cursor/rules/mvu角色卡.mdc`
references/script-analysis.md
- Network accessline 47
curl -sL '<URL>' | head -100
scripts/extract_card.py
scripts/get_entry.py
- Executable scriptline 1
#!/usr/bin/env python3
scripts/list_entries.py
- Executable scriptline 1
#!/usr/bin/env python3
scripts/start.sh
- Executable scriptline 1
#!/usr/bin/env bash - Network accessline 6
if ! command -v pi &>/dev/null; then echo "错误: pi 未安装,请先安装 pi coding agent" >&2 echo "安装指引: https://github.com/earendil-works/pi-coding-agent" >&2 - Network accessline 7
echo "错误: pi 未安装,请先安装 pi coding agent" >&2 echo "安装指引: https://github.com/earendil-works/pi-coding-agent" >&2 exit 1 - Network accessline 8
echo "安装指引: https://github.com/earendil-works/pi-coding-agent" >&2
How the signals fit together
Facts from a static read of the files at the pinned commit. They describe what the code does, not why.
Signal confidence: 19 facts
- Network access in prose of documentationREADME.md:34Low confidence
- Network access in code of documentationREADME.md:43Low confidence
- Network access in prose of documentationREADME.md:126Low confidence
- Network access in prose of documentationREADME_ZH.md:34Low confidence
- Network access in code of documentationREADME_ZH.md:43Low confidence
- Network access in prose of documentationREADME_ZH.md:126Low confidence
- Broad permissions in the frontmatter of the SKILL.mdSKILL.md:1High confidence
- Network access in prose of documentationreferences/mvu-mapping.md:31Low confidence
- Network access in prose of documentationreferences/mvu-mapping.md:32Low confidence
- Network access in code of documentationreferences/script-analysis.md:47Low confidence
- Executable script in code of a file that runsscripts/extract_card.py:1High confidence
- Possible obfuscation in code of a file that runsscripts/extract_card.py:82High confidence
- Possible obfuscation in code of a file that runsscripts/extract_card.py:113High confidence
- Executable script in code of a file that runsscripts/get_entry.py:1High confidence
- Executable script in code of a file that runsscripts/list_entries.py:1High confidence
- Executable script in code of a file that runsscripts/start.sh:1High confidence
- Network access in code of a file that runsscripts/start.sh:6High confidence
- Network access in code of a file that runsscripts/start.sh:7High confidence
- Network access in code of a file that runsscripts/start.sh:8High confidence
Works with
| Agent | Status | Evidence | Install to |
|---|---|---|---|
| Claude Code | Compatible | Standard SKILL.md SKILL.md | .claude/skills/<skill-name>/ |
| Codex | Compatible | Standard SKILL.md SKILL.md | .agents/skills/<skill-name>/ |
| Cursor | Compatible | Standard SKILL.md SKILL.md | .cursor/skills/<skill-name>/ |
| Gemini CLI | Compatible | Standard SKILL.md SKILL.md | .gemini/skills/<skill-name>/ |
| GitHub Copilot | Compatible | Standard SKILL.md SKILL.md | .github/skills/<skill-name>/ |
| OpenCode | Compatible | Standard SKILL.md SKILL.md | .opencode/skills/<skill-name>/ |
- Native The repository ships a manifest, folder or frontmatter field for this agent.
- Compatible A standard SKILL.md that this agent can read.
- Unknown No evidence either way.
Derived automatically from paths, manifests and frontmatter. Not tested on each agent.