neo-clarification
Skill7 starsChanged 6 months agoMITView on GitHub
Use this skill when the user gives vague, emotional, fragmented, screenshot-based, or complaint-style requirements and wants them converted into structured specs, acceptance criteria, or clarifying questions.
In this repository28 other files it gives its agents.
- AGENTS.mdAGENTS.md
- generate-skillSkill
- neo-azure-pipelinesSkill
- neo-clean-architectureSkill
- neo-code-reviewSkill
- neo-csharp-interface-generatorSkill
- neo-csharpSkill
- neo-dotnet-ef-coreSkill
- neo-dotnet-minimal-apisSkill
- neo-dotnet-mvcSkill
- neo-dotnetSkill
- neo-dotnet-tag-helperSkill
- neo-dotnet-webapiSkill
- neo-harness-projectSkill
- neo-harnessSkill
- neo-javascriptSkill
- neo-opentelemetrySkill
- neo-prSkill
- neo-python-managerSkill
- neo-pythonSkill
- neo-rabbitmqSkill
- neo-rustSkill
- neo-stop-slopSkill
- neo-sub-agentSkill
- neo-swiftSkill
- neo-swift-uiSkill
- neo-typescriptSkill
- neo-vueSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 48
Read the standard markdown structure from: clarification-template.md
- Distinctive: line 54
User Story: Use the strict format: "身為...
- Distinctive: line 70
Strictly No Guesswork: Do NOT invent features that the user did not hint at.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Requirement Clarification Specifications
Apply the Inversion & Generator Pattern. Follow this protocol strictly to translate raw, chaotic user complaints and screenshots into clean, structured system specifications.
1. Perceive Phase
- Information Extraction:
- Carefully read the user's text and inspect any attached screenshots or logs.
- Filter out emotional noise, frustrations, and blame.
- Separate objective facts (what is currently happening or visible) from user expectations (what they wanted to accomplish).
- Identify System Boundaries:
- Determine the scope, domain, and potential technical layers affected by the feedback (e.g., frontend rendering, network APIs, database states, permission groups).
2. Reason Phase
- Load Analysis Framework:
- Always read the external analysis guide before starting your deduction: 5w1h-framework.md
- Context Reconstruction (5W1H):
- Map the extracted facts to the 5W1H framework (Who, Where, When, What, Why, How).
- Formulate logical hypotheses on the root causes of UI anomalies or system behaviors.
- Identify Gaps:
- Pinpoint critical missing information (e.g., browser environment, specific action steps, parameters, error logs).
- Prepare a list of clarifying questions to ask the user.
3. Act Phase
Generate a structured "Requirement Translation and Clarification Report" strictly in Traditional Chinese (Taiwan). Follow these steps:
- Load Output Template:
- Read the standard markdown structure from: clarification-template.md
- Compile the Report:
- Fill in the template using Traditional Chinese.
- Context Restoration: Present objective facts concisely without emotional adjectives.
- User Story: Use the strict format: "身為... 我想要... 以便於..."
- System Requirements & Hypotheses: Highlight key rendering, API, and validation checkpoints for the development team.
- Open Questions: List between 2 and 10 polite, precise, and constructive clarifying questions.
- Self-Validation:
- Proactively validate your report using the non-interactive python script:
(Or verify mentally that all sections exist and the number of questions is between 2 and 10 before final output).uv run skills/neo-clarification/scripts/validate-requirements.py -i <path_to_saved_report>
- Proactively validate your report using the non-interactive python script:
4. Communication Guidelines
- Maintain Empathetic Neutrality: Acknowledge the user's difficulty, but never agree that the system is "broken" or "a disaster" in the official report. Use neutral, objective descriptions.
- Strictly No Guesswork: Do NOT invent features that the user did not hint at. Ask clarifying questions instead.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

