图像同步
- 作者仓库星标 0
- 作者更新于 实时读取
- 作者仓库 skills-registry
- 领域
- 设计与多媒体
- 兼容 Agent
-
- Claude Code
- Cursor
- Cline
- Codex
- Windsurf
- Gemini CLI
- +20
- 信任分
- 88 / 100 · 社区维护
- 作者 / 版本 / 许可
- @tomevault-io · 未声明 license
- Token 消耗评级
- 低消耗
- 接入复杂程度
- 即装即用
- 是否需要外部 API Key
- 不需要
- 兼容的系统
- 未声明(默认跨平台)
- 底层运行要求
- 无特殊要求
- 文件与系统权限
-
- 只读
- 允许写入 / 修改
- 读取环境变量
- 网络行为
- 仅限本地
- 安装命令数
- 26 条
档案由构建时根据 SKILL.md 与安装命令自动衍生,可能与作者实际意图存在差异。
需要注意: 未限定 allowed-tools,默认拥有全部工具权限。
---
name: repo-guidance-sync
description: Keep repository documentation, AGENTS.md, and contributor/test guidance aligned with code, desig…
category: 设计与多媒体
runtime: 无特殊运行时
---
# repo-guidance-sync 输出预览
## PART A: 任务判断
- 适用问题:视觉内容、演示材料、信息图或设计交付。
- 输入要求:目标材料、限制条件、期望输出和验收方式。
- 证据边界:围绕“Overview / Workflow / AGENTS.md Rules”读取原文规则,不把推断写成作者承诺。
## PART B: 执行结果
- **01** 任务判断:确认你的需求是否属于视觉内容、演示材料、信息图或设计交付,并标出输入、限制和预期结果。
- **02** 执行计划:优先按“Overview / Workflow / AGENTS.md Rules”拆成步骤,说明每一步会读取什么、修改什么、产出什么。
- **03** 交付结果:给出可复制的命令、文件改动、检查清单或内容草稿,并说明如何继续迭代。
- **04** 风险边界:结合 读取文件、写入/修改文件、读取环境变量、主要在本地完成、通常不需要额外 API Key 给出执行前确认项。
## Running Rules
- 读取文件、写入/修改文件、读取环境变量;主要在本地完成;通常不需要额外 API Key。
- 先小样例验证,再放大到真实任务。
- 交付时同时给结果、检查口径和下一步迭代建议。 原文没有稳定的斜杠命令要求。安装验证后通常全局生效,直接在对话里点名这个 Skill 并描述任务即可。
告诉 Agent 目标文件或材料、期望结果、不可改范围、是否允许联网或执行命令。本 Skill 的权限画像是:读取文件、写入/修改文件、读取环境变量。
先用一个小任务确认它会围绕“Overview / Workflow / AGENTS.md Rules”工作;涉及文件或命令时,先看 diff、日志、预览或测试结果。
检查最终产物是否包含明确结果、必要证据和下一步动作;如果输出泛泛而谈,就补充输入、边界和验收标准后重跑。
---
name: repo-guidance-sync
description: Keep repository documentation, AGENTS.md, and contributor/test guidance aligned with code, desig…
category: 设计与多媒体
source: tomevault-io/skills-registry
---
# repo-guidance-sync
## 什么时候使用
- 用于审阅代码、文档或方案并给出可执行反馈 适合处理界面、视觉、封面、信息图或演示材料交付,核心价值是把输入、判断、执行、验证和交付边界固定下来,避免 Agent 泛泛回答。 把任务拆成可执行、可检查、可继续迭代的步骤;通常不需要额外 A…
- 面向视觉内容、演示材料、信息图或设计交付,优先处理能明确输入、步骤和验收标准的工作。
## 需要提供什么
- 目标材料、目录范围、期望结果和不可改动内容。
- 是否允许联网、执行命令、读写文件或调用外部服务。
## 执行规则
- 围绕「Overview / Workflow / AGENTS.md Rules」组织步骤,不把推断写成作者事实。
- 读取文件、写入/修改文件、读取环境变量;主要在本地完成;通常不需要额外 API Key。
- 先跑小样例,确认结果可检查后再扩大任务范围。
## 输出要求
- 给出最终产物、关键证据、验证方式和下一步动作。
- 信息不足时标记 unknown,不编造命令、平台或依赖。 作者原文负责流程事实;仓库文件负责来源和命令;流狐只补充适用场景、限制和质量判断。
skill "repo-guidance-sync" {
输入层 -> 用户目标 + 目标文件 + 禁止范围 + 验收标准
上下文层 -> Overview / Workflow / AGENTS.md Rules
规则层 -> SKILL.md 触发条件 / 执行顺序 / 输出格式
运行层 -> 无特殊运行时 | 读取文件、写入/修改文件、读取环境变量 | 主要在本地完成
安全层 -> 通常不需要额外 API Key + 小任务验证 + diff / 日志复核
输出层 -> 可复制结果 + 检查清单 + 下一步迭代
} Repo Guidance Sync
Overview
Use this skill to decide what documentation must change after a code, design, or product update, then make those edits in the right files with verifiable links back to source behavior.
Workflow
Identify the change surface.
- Inspect the diff, touched packages, new commands, config flags, UI behavior, API shapes, CRDs, generated schemas, deployment manifests, and tests.
- If the change affects setup, CLI workflows, operator reconciliation, ingress/gateway behavior, auth/policy, registry/image handling, observability, or cache-mode assumptions, identify the matching E2E scenario in
test/e2e/kind.shor adjacent integration tests before editing docs. - Search for existing docs before adding new files:
rg -n "<command|field|feature|concept>" README.md AGENTS.md docs website config api internal services. - Prefer updating the nearest existing guide, reference, or runbook over creating a new document.
Classify required doc updates.
- User-facing product change: update README, docs guides, CLI examples, API/reference pages, screenshots or UI copy references if present.
- Developer workflow change: update AGENTS.md, build/test instructions, local setup notes, failure modes, environment variables, or repo map entries.
- Codebase architecture or contributor-flow change: review
docs/internals/and decide whether the overview, package tours, or Mermaid diagrams need updates. Add or revise diagrams only when they clarify package boundaries, request flow, reconciliation flow, service interactions, or contributor decision points. - Design change: update design-system rules, component guidance, UX behavior notes, or docs that describe screens/workflows.
- Operational change: update install, deployment, TLS, registry, Kubernetes, observability, and rollback/debug instructions.
- Schema/contract change: update CRD/API docs, examples, generated docs if the repo owns them, and any golden snapshots tied to CLI help.
- E2E-visible behavior change: update or add the relevant E2E assertion in
test/e2e/kind.shwhen the behavior is exercised through Kind, real MCP traffic, policy/session state, OAuth, observability, registry/image paths, orcluster doctor. If narrower unit/integration coverage is sufficient, note why in the final validation summary.
Verify source truth before writing.
- Treat code, tests, CRD types, OpenAPI specs, CLI definitions, workflow files, and manifests as source of truth.
- For commands and examples, confirm exact flags, defaults, resource names, paths, and environment variables from implementation.
- Avoid inventing roadmap promises or behavior not present in code unless the user explicitly asks for planned-product wording.
Edit with the right scope.
- Keep docs concise and task-oriented; remove stale claims instead of layering caveats.
- Preserve the repo's voice, terminology, command style, and existing document structure.
- For
docs/internals/, prefer updating the nearest existing walkthrough, mental model, or diagram over adding a new page. Use diagrams for relationships and flows, not for listing files that a table can express more clearly. - Update AGENTS.md only for durable contributor or agent guidance, not one-off implementation details.
- Keep examples copy-pasteable: include required tags, namespaces, paths, auth headers, and prerequisite environment variables.
- Update links and table-of-contents entries when adding or renaming pages.
Validate.
- Run the narrowest relevant checks: markdown/docs build if available, golden tests for CLI help changes, generated-doc drift checks when docs are generated, E2E scenario checks for behavior covered by
test/e2e/kind.sh, and targeted code tests if examples exercise behavior. - At minimum, run searches for old field names, commands, or stale wording that the change replaces.
- If validation cannot be run, state the blocker and what was manually checked.
- Run the narrowest relevant checks: markdown/docs build if available, golden tests for CLI help changes, generated-doc drift checks when docs are generated, E2E scenario checks for behavior covered by
AGENTS.md Rules
Update AGENTS.md when a change affects how future agents or contributors should work in the repo. Good reasons include new required checks, changed setup commands, new failure modes, changed repository layout, new service ownership, changed CLI workflows, or new operational safety rules.
Do not put product marketing, exhaustive API reference, release notes, or transient PR context in AGENTS.md. Link or summarize stable docs instead.
Output Standard
When finished, summarize:
- docs and AGENTS.md files changed
- source behavior each update reflects
- validation run, including searches for stale terms
- known doc gaps left intentionally open
Source: Agent-Hellboy/mcp-runtime — distributed by TomeVault.
先判断是否适合
作者设计意图
作者的方法与取舍
边界和复核