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 agentmods add skills/thestack-ai/swing-skills/swing-clarifynpx skills add TheStack-ai/swing-skills --skill swing-clarifygit clone --depth 1 https://github.com/TheStack-ai/swing-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/thestack-ai/swing-skills/swing-clarify)<a href="https://agentmods.dev/skills/thestack-ai/swing-skills/swing-clarify"><img src="https://agentmods.dev/badge/skills/thestack-ai/swing-skills/swing-clarify.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00107 | $0.02299 |
| Opus 5 | $0.00053 | $0.01149 |
| Sonnet 5 | $0.00021 | $0.00460 |
| Haiku 4.5 | $0.00011 | $0.00230 |
Grade A, and why
swing-clarify 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 6d 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 — 205 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Scope Clarifier
Prevents the most common AI failure: rushing to execute before understanding what's actually needed.
Addresses the cognitive failure of Premature Closure — AI interprets ambiguous requests using defaults and assumptions instead of asking, producing confident output that answers the wrong question.
Rules (Absolute)
- Never execute before clarifying. If ambiguity score is above threshold, generate questions FIRST. Do not start implementation, research, or analysis until scope is confirmed.
- Maximum 3 questions. Respect the user's time. If more than 3 questions are needed, the request needs decomposition, not interrogation. Ask the 3 highest-impact questions.
- Questions must be actionable. Every question must change what you build. "What's your timeline?" is only valid if it affects scope. "Should this handle authentication?" is always valid if auth wasn't mentioned.
- Prefer multiple choice over open-ended. "Should auth use (a) session cookies, (b) JWT, or (c) OAuth2 with a provider?" beats "How should auth work?"
- State your default assumption. For each question, state what you WOULD assume if the user doesn't answer. This lets them skip questions where the default is fine.
- Clear requests get a green light, not questions. If the request is unambiguous, say so and proceed. Do not ask questions for the sake of asking.
- Never block on style preferences. Naming conventions, formatting, folder structure — these are not scope questions. Use project conventions or sensible defaults.
Process
Stage 1: 5W1H Decomposition
Break the request into six dimensions:
| Dimension | Question | Example Gap |
|---|---|---|
| What | What exactly is being built/changed? | "Build auth" — login? signup? password reset? SSO? |
| Who | Who uses this? What roles/permissions? | "Users can edit" — all users? admins only? owners? |
| Where | Where does this live? What system/service? | "Add to the API" — which API? new endpoint? existing? |
| When | What triggers this? What's the lifecycle? | "Send notifications" — real-time? batched? on what event? |
| Why | What problem does this solve? What's the success criteria? | "Improve performance" — latency? throughput? cost? |
| How | Are there constraints on implementation? | "Use the existing stack" — which parts? any exceptions? |
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.
- 6d ago First seen · 205 lines · 107 tokens per session scan A 7a020ee68cb2
swing-clarify is a skill published in the GitHub repository TheStack-ai/swing-skills (40 stars, last pushed 4mo ago), licensed MIT. It adds 107 tokens to every session and 2,299 once invoked, about $0.0005 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
prompting-buddy
Generate, review, and optimize project-level prompt files for Claude Code -- including CLAUDE.md, .claude/rules/.md, and .claude/agents/.md. Use this skill whenever someone mentions CLAUDE.md, AGENTS.md, project prompts, coding instructions, agent definitions, rules files, or asks for help setting up Claude Code for a…
prompt-master
Generates optimized prompts for AI tools. Activates only when the user explicitly asks to write, fix, improve, or adapt a prompt for a specific AI tool (LLM, Cursor, Midjourney, image AI, video AI, coding agents, etc.). Does not activate for general conversation, coding tasks, document writing, or other…
skill-curator
中文 Skill 收录评估器。用于核验 GitHub 上的 Claude Code、Codex、Agent Skills、Plugin、MCP、CLI 或 Agent 工作流是否值得加入精选列表；检查真实资产、安装方式、活跃度、重复项和安全边界，并生成分类、中文描述与可追溯证据。.
git-workflow
安全处理 Git 状态检查、提交信息、commit、分支、push、PR 和 rebase。用于用户要求检查改动、生成或创建提交、管理分支、推送、发起 PR 或整理历史时；严格区分每个动作的授权，并保护工作树中已有和无关的修改。.
i18n-helper
国际化/本地化助手 — 扫描代码中的硬编码文本、生成 i18n 配置、批量翻译.
log-analyzer
日志分析助手 — 智能解析日志文件，识别异常模式，定位问题根因.