clarify-requirements
Use for unresolved product choices or explicit requirements discovery—not discoverable facts, implementation details, approvals, or final goals.
Skill by shiyuan17
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/shiyuan17-clarify-requirements
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/clarify-requirements
What's inside
- Skill name
clarify-requirements- Description
- Use for unresolved product choices or explicit requirements discovery—not discoverable facts, implementation details, approvals, or final goals.
- License
- MIT
- Instructions
- 32 lines in SKILL.md · view on GitHub
Files (4)
- SKILL.md
- openai.yaml · agents/
- metadata.json
- examples.md · references/
SKILL.md (first lines)
# 澄清关键需求 先消除事实未知,再按请求深度处理真正改变结果的产品决定。 ## 普通解阻 1. 检查仓库、测试、文档和公开契约;不要向用户询问可发现事实。 2. 将剩余未知分为:安全审批、阻塞产品决定、可逆实现选择。 3. 安全审批先核对已有授权的目标、环境、动作与范围;生产、凭据、外部写入或不可逆操作缺授权或新增风险时,只暂停受影响操作并请求批准,已有覆盖授权不重复确认。可逆实现选择沿用仓库惯例或采用最小可逆默认值。同步归属:`docs/rules/governance-core.md`。 4. 仅当剩余分歧无法从当前意图、事实和项目惯例解决,且不同答案实质改变产品结果、数据语义、安全或验收时,才暂停依赖该答案的工作;等待回答时继续独立的已授权工作。 5. 候选超过三个时按影响×不确定性排序,保留其他重要问题待解;同轮合并最多三个互不依赖的决定,依赖前置答案的留到下一轮。 6. 每题给出推荐项、二至三个互斥选择和一句结果影响。 7. 对无法直接发现且可能改变方案的未知项进行信息价值分析:记录未知项、可能改变的决定、最低成本验证、成功标准、失败处理和停止条件;优先验证最可能改变结果的未知,而不是先扩大调研范围。用具体输入与预期结果暴露真正改变验收的分歧;复用用户已明确的例子和保留行为,不为清晰小改动重新采访或创建规格。 8. 回答关闭分支后立即继续,不追加通用设计、计划或最终确认。对照示例见 [references/examples.md](references/examples.md)。 ## 显式需求发现 用户明确要求采访、探索或分析需求时,补齐:目标用户、真实问题、期望结果、当前基线、范围与非目标、成功证据、假设与风险。优先用仓库事实覆盖维度,只询问剩余高影响决定;对仍会改变结果的未知项提出最小可行实验,并写明成功、失败和停止条件;把用户指定的技术手段记录为待验证假设,不替代结果目标。 非平凡特性可输出自包含单页规格:命名涉及的文件或接口、显式非目标、以端到端验证步骤收尾;一句话能描述的小改动跳过规格。默认输出紧凑的需求决定摘要;除非用户要求,不创建规格文档。若用户要求最终 `/goal` 或目标书,交由目标定义能力独立处理。 ## 自治姿态 读取项目配置 `clarification.posture`:action-leaning 偏向采用最小可逆默认值直接推进,仅在不可逆或高影响分支阻塞;balanced(默认)按上文判断;conservative 对尚未解决的高影响分歧更谨慎。配置缺失视为 balanced;三种姿态都不为清晰请求、可发现事实或已关闭决定制造问题,跨模块或公共契约本身不是确认触发器。 ## 禁止 - 不询问文件位置、实现方法、工具选择或验证命令。 - 不把生产、凭据、红区或不可逆动作包装成普通产品问题。 - 不为清晰任务制造选项,也不静默决定高影响产品分支。 ## 正反例 - 该问:「导出的报告要 PDF 还是 Excel」——改变交付物与验收的产品选择,仓库里没有答案。 - 不该问:「入口函数在哪个文件」(可发现事实)、「用哪个测试框架」(实现细节,按仓库惯例)、「可以改这个文件吗」(已在授权范围内)。 - 需求已明确:可执行程度足够时直接推进,不追加确认;只有不可逆或高影响分支才停顿。
From shiyuan17/Vibe-Harness, MIT.
More from shiyuan17/Vibe-Harness
- 0.570score
agentmemory
shiyuan17
Use when saving, recalling, resuming, forgetting, summarizing, or listing Agentmemory project memories and sessions, including commit history and commit…
SkillAFreshness: AClaude Code +9 - 0.570score
stale-cleanup
shiyuan17
Use when cleaning dead code, stale references, outdated docs, orphan resources, stale memory, or stale indexes; producing evidence-based removal lists; or…
SkillAFreshness: AClaude Code +9 - 0.565score
systematic-debugging
shiyuan17
Use when a specific failure, error, or flaky test needs its root cause proven before fixing—not for applying known fixes.
SkillAFreshness: AClaude Code +9 - 0.564score
define-goal
shiyuan17
Use to write, refine, or activate `/goal` or execution/exploration objectives—not goal status operations.
SkillAFreshness: AClaude Code +9 - 0.564score
frontend-design
shiyuan17
Use for frontend visual direction, polish, responsiveness, or redesign—not logic-only work.
SkillAFreshness: AClaude Code +9 - 0.561score
eval-driven-development
shiyuan17
Use for nondeterministic Agent rules, Skills, prompts, Hooks, or routing—not deterministic logic, bug fixes with known causes, or deterministic refactors.
SkillAFreshness: AClaude Code +9 - 0.561score
vb-git-develop-flow
shiyuan17
Use only when explicitly requested as Vb Git Develop Flow, $vb-git-develop-flow, or legacy gd-flow to deliver through an isolated worktree into develop; never…
SkillAFreshness: AClaude Code +9 - 0.561score
vb-git-split-push
shiyuan17
Use only when explicitly requested as Vb Git Split Push, $vb-git-split-push, or legacy $git-deliver to commit task changes in logical groups and push the…
SkillAFreshness: AClaude Code +9 - 0.559score
bug-finding
shiyuan17
Use when reviewing existing code for latent logic, flow, feature, state, consistency, requirement, or API bugs with no failing symptom—read-only audit, not…
SkillAFreshness: AClaude Code +9 - 0.558score
api-and-interface-design
shiyuan17
Use for public APIs, schemas, events, config, or consumed module boundaries.
SkillAFreshness: AClaude Code +9 - 0.558score
task-decomposition
shiyuan17
Use to split implementation plans into a Goal, Task DAG, and prompts for verifiable work.
SkillAFreshness: AClaude Code +9 - 0.553score
runtime-cross-repo-rollout
shiyuan17
Use for coordinated contracts and rollout across repositories—not modules in one repo.
SkillAFreshness: AClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Oct 14, 2026 (15 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/shiyuan17/Vibe-Harness
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
Same skill elsewhere
Copies of this SKILL.md published by others; search shows only one of them.
More from shiyuan17
- 0.570score
agentmemory
shiyuan17
Use when saving, recalling, resuming, forgetting, summarizing, or listing Agentmemory project memories and sessions, including commit history and commit…
SkillAFreshness: AClaude Code +9 - 0.570score
stale-cleanup
shiyuan17
Use when cleaning dead code, stale references, outdated docs, orphan resources, stale memory, or stale indexes; producing evidence-based removal lists; or…
SkillAFreshness: AClaude Code +9 - 0.565score
systematic-debugging
shiyuan17
Use when a specific failure, error, or flaky test needs its root cause proven before fixing—not for applying known fixes.
SkillAFreshness: AClaude Code +9 - 0.564score
frontend-design
shiyuan17
Use for frontend visual direction, polish, responsiveness, or redesign—not logic-only work.
SkillAFreshness: AClaude Code +9 - 0.564score
define-goal
shiyuan17
Use to write, refine, or activate `/goal` or execution/exploration objectives—not goal status operations.
SkillAFreshness: AClaude Code +9 - 0.561score
eval-driven-development
shiyuan17
Use for nondeterministic Agent rules, Skills, prompts, Hooks, or routing—not deterministic logic, bug fixes with known causes, or deterministic refactors.
SkillAFreshness: AClaude Code +9
Cite this page
clarify-requirements skill. Agent Mart. https://agentmart.dev/skills/shiyuan17-clarify-requirements (last updated 2026-10-24)