Treasoni/Study-Notes

research-collector

Collect and curate technical research in two gated stages with compact source records. Use for systematic research, source gathering, research materials, or when a learning-note workflow is at collection phase P1 or P2.

原文语言：英语

tool-discovery

发现当前运行时可用于资料收集的网页、浏览器、MCP 与研究技能，并刷新项目资料工具指南。用于询问可用工具、搜索工具、研究能力或资料收集方案时。

note-beautifier

Obsidian 笔记智能美化与发布。用于将最终学习笔记处理成 Obsidian Markdown，补 frontmatter、标签、Callout、双链，并保存到用户指定位置。触发词：美化、Obsidian、优化格式、笔记美化、发布到 vault、beautify。

research-collector

Collect and curate technical research in two gated stages with compact source records. Use for systematic research, source gathering, research materials, or when a learning-note workflow is at collection phase P1 or P2.

原文语言：英语

research-planner

学习笔记需求澄清与引导。分析用户学习需求，引导明确学习目标和方向，调用 workflow-orchestrator 生成项目结构。触发词：想学、帮我整理、研究一下、了解一下、不知道从哪开始、research planning、explore topic。

tool-discovery

发现当前运行时可用于资料收集的网页、浏览器、MCP 与研究技能，并刷新项目资料工具指南。用于询问可用工具、搜索工具、研究能力或资料收集方案时。

git-guardrails-claude-code

Set up Claude Code hooks to block dangerous git commands (push, reset --hard, clean, branch -D, etc.) before they execute. Use when user wants to prevent destructive git operations, add git safety hooks, or block git push/reset in Claude Code.

原文语言：英语

writing-for-agents

Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.

原文语言：英语

batch-note-updater

多篇既有学习笔记的批量更新编排。用于用户想一次更新一个目录、文件列表、Obsidian vault 子目录或多篇旧笔记，例如“批量更新这些笔记”“多篇笔记过时了”“把一组笔记更新到新版本”“refresh multiple notes”。先生成更新清单和批量计划，经用户确认后逐篇调用 note-updater 局部 patch，避免全文重写和批量误覆盖。

code-review

Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes — Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating issue/spec asked for?). Runs both reviews…

原文语言：英语

codebase-design

Shared vocabulary for designing deep modules. Use when the user wants to design or improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable or AI-navigable, or when another skill needs the deep-module…

原文语言：英语

diagnosing-bugs

Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.

原文语言：英语

digest

自我学习阶段。回顾本次会话，记录真实发生的学习点和错误到 .learnings/；当经验库过长时压缩去重并更新 RULES.md；如果发现重复错误或规则失效，转交 maintain-learnings 先修源头。用户明确要求记录学习、复盘、写入 learnings、digest 时触发。

domain-modeling

Build and sharpen a project's domain model. Use when discussing codebase terminology, writing or editing a CONTEXT.md, or recording or editing an ADR.

原文语言：英语

improve-codebase-architecture

Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.

原文语言：英语

legacy-note-importer

旧笔记批量导入、盘点和规范化。用于用户已经有一堆 Markdown/Obsidian/零散学习笔记，想接入本项目工作区、按项目 Obsidian 规范补 frontmatter、标签、双链、Callout、MOC，并保留原始文件时使用。触发词：旧笔记导入、已有笔记、一堆笔记、批量整理、迁移到这个项目、按项目规范、import existing notes、normalize notes。

maintain-learnings

维护 .learnings/ 经验库，把过多或反复出现的学习记录、错误日志、规则失效问题聚类诊断，追溯并修改对应 skill、模板、hook、校验脚本或项目规则；修复并验证后再归档或移除已解决记录。用户提到 learnings 太多、错误反复犯、清理经验库、维护自我学习、压缩错误日志或从错误中修技能时触发。

moc-organizer

为 Obsidian 生成或更新 MOC（Map of Content）目录笔记。用于用户要求“生成 MOC”“整理目录”“把新笔记加入目录”“每次加入笔记自动更新索引”等场景。只维护索引、分组、双链和简短说明，不复制正文。

note-beautifier

Obsidian 笔记智能美化与发布。用于将最终学习笔记处理成 Obsidian Markdown，补 frontmatter、标签、Callout、双链，并保存到用户指定位置。触发词：美化、Obsidian、优化格式、笔记美化、发布到 vault、beautify。

note-starter

启动新主题学习笔记。用于用户说“开始写笔记”“启动写笔记”“创建学习笔记”或明确想为新主题建立学习笔记时。检查可恢复运行后，委派给 research-planner 初始化 learning-note-flow；不用于旧笔记导入或更新。

note-updater

更新过时的既有学习笔记。用于用户说“更新这篇笔记”“这篇笔记过时了”“根据新资料刷新旧笔记”“同步到 Obsidian 旧笔记”等场景。先定位旧笔记、判断过时段落、收集最小必要新资料，再局部 patch，避免重写整篇和浪费 token。

prompt-cache-optimizer

审计并优化 LLM 提示缓存命中率、输入 token、延迟与调用成本。用于用户要求“优化缓存命中”“降低 token 成本”“审计 LLM 调用”“提示词缓存优化”“优化 AI 调用费用”，或需要为可自动读取 API 调用指标的应用建立可观测性和固定回归样本时。

原文语言：多语言混合

research-planner

学习笔记需求澄清与引导。分析用户学习需求，引导明确学习目标和方向，调用 workflow-orchestrator 生成项目结构。触发词：想学、帮我整理、研究一下、了解一下、不知道从哪开始、research planning、explore topic。

security-secret-audit

Audit a Git repository for exposed API keys, tokens, passwords, private keys, project-owned source/configuration risks, and sensitive-file policy gaps without printing secret values. Use when asked to check project security, scan for API leaks, review files…

原文语言：英语

skill-creator

Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or…

原文语言：英语

sync-skill-registry

技能注册表同步工具。扫描任意 agent skill 目录中的 */SKILL.md 并自动更新对应 skill-invocation.md 中的技能列表表格。用于新增、修改或删除本地技能后同步注册表。触发词：同步注册表、更新技能列表、sync skill registry、update skill registration、刷新技能列表、同步技能表格。

wizard

Generate an interactive bash wizard that walks a human through steps only they can perform. Use when provisioning infrastructure, setting up credentials or CI secrets, walking an unfamiliar third-party dashboard, or running a one-off migration or cutover.…

原文语言：英语

workflow-orchestrator

业务工作流实例化器。由 planner 技能调用，接收 workflow_id、topic、project_slug 等参数，按 .codex/workflows/{workflow-id}/state-template.md 生成 workspace/workflow-runs/{run-id}.workflow.md。不直接面向用户；通用状态格式和状态流转由 workflow-todo-state 负责。

workflow-todo-state

Create or retrofit reusable named workflow state machines for multi-step agent projects. Use when a project needs recoverable workflow state files, multiple workflow definitions, phase gates, restart-safe agent workflows, explicit skipped/blocked states,…

原文语言：英语

prompt-cache-optimizer

审计并优化 LLM 提示缓存命中率、输入 token、延迟与调用成本。用于用户要求“优化缓存命中”“降低 token 成本”“审计 LLM 调用”“提示词缓存优化”“优化 AI 调用费用”，或需要为可自动读取 API 调用指标的应用建立可观测性和固定回归样本时。

原文语言：多语言混合

manifest-platform

Install, configure, migrate, and validate a portable manifest registry for agent workflows, skills, subagents, and hooks. Use when a project needs manifest.yaml-based discovery, artifact versioning, dependency checks, capability declarations,…

原文语言：英语

manifest-platform

Install, configure, migrate, and validate a portable manifest registry for agent workflows, skills, subagents, and hooks. Use when a project needs manifest.yaml-based discovery, artifact versioning, dependency checks, capability declarations,…

原文语言：英语

workflow-todo-state

Create or retrofit reusable named workflow state machines for multi-step agent projects. Use when a project needs recoverable workflow state files, multiple workflow definitions, phase gates, restart-safe agent workflows, explicit skipped/blocked states,…

原文语言：英语

code-review

Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes — Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating issue/spec asked for?). Runs both reviews…

原文语言：英语

codebase-design

Shared vocabulary for designing deep modules. Use when the user wants to design or improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable or AI-navigable, or when another skill needs the deep-module…

原文语言：英语

diagnosing-bugs

Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.

原文语言：英语

domain-modeling

Build and sharpen a project's domain model. Use when the user wants to pin down domain terminology or a ubiquitous language, record an architectural decision, or when another skill needs to maintain the domain model.

原文语言：英语

git-guardrails-claude-code

Set up Codex hooks to block dangerous git commands (push, reset --hard, clean, branch -D, etc.) before they execute. Use when user wants to prevent destructive git operations, add git safety hooks, or block git push/reset in Codex.

原文语言：英语

improve-codebase-architecture

Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.

原文语言：英语

maintain-learnings

维护 .learnings/ 经验库，把过多或反复出现的学习记录、错误日志、规则失效问题聚类诊断，追溯并修改对应 skill、模板、hook、校验脚本或项目规则；修复并验证后再归档或移除已解决记录。用户提到 learnings 太多、错误反复犯、清理经验库、维护自我学习、压缩错误日志或从错误中修技能时触发。