clarify
Skill98 starsChanged 9 months agoMITView on GitHub
Identify and resolve ambiguities in feature specifications. Use when spec has unclear requirements. Triggers on: clarify spec, resolve ambiguities, clarify requirements.
We split this skill into the rules it gives an agent
8 rules. 0 are common, 8 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 78 edited versions we read keep these.
8
Distinctive
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 27
When clarifying requirements, ensure answers are:
- Distinctive: line 49
Update spec with answers
- Distinctive: line 62
Check for these 10 types:
- Distinctive: line 121
If routing preference is unclear or missing from spec, ask:
- Distinctive: line 157
Update spec.md with clarifications
- Distinctive: line 159
Add concrete details based on answers
- Distinctive: line 231
Update spec immediately after clarification
- Distinctive: line 232
Keep clarification log for future reference
“Common” is measured: kept by at least 6 in 10 of the 78 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Specification Clarification Tool
Systematically identify and resolve ambiguities in feature specifications.
SpecKit Workflow
Clarify can be invoked at any stage when ambiguities arise:
specify ↔ clarify ↔ plan ↔ clarify ↔ tasks
The skill helps resolve ambiguities that could affect:
- Test design (unclear acceptance criteria)
- Routing decisions (unclear complexity)
- Implementation choices (unclear requirements)
Testability Requirements
When clarifying requirements, ensure answers are:
- Specific enough to write tests against
- Measurable (can verify with automated tests)
- Unambiguous (no multiple interpretations)
Good vs Bad Clarifications
| Bad (vague) | Good (testable) |
| "How should validation work?" | "Should email validation reject 'user@localhost'?" |
| "What about error handling?" | "Should invalid password show 'Invalid credentials' or 'Password too short'?" |
| "How secure should it be?" | "Should we require 2FA for admin users?" |
Every clarification should result in a testable requirement.
The Job
- Read spec.md
- Scan for 10 types of ambiguities (including testability)
- Present clarification questions to user
- Update spec with answers
- Save clarification log
Step 1: Load Specification
Read relentless/features/NNN-feature/spec.md
Step 2: Scan for Ambiguities
Check for these 10 types:
1. Behavioral Ambiguities
- What happens when [action]?
- How should system respond to [event]?
- Example: "What happens when user enters wrong password 5 times?"
2. Data Ambiguities
- What are required vs optional fields?
- What validation rules apply?
- Example: "What email format validation should we use?"
3. UI/UX Ambiguities
- Where should this be displayed?
- How should user interact?
- What visual style?
- Example: "Should password field show/hide toggle?"
4. Integration Ambiguities
- Which external API/service?
- What happens if service unavailable?
- Example: "Which email provider: SendGrid, Mailgun, or AWS SES?"
5. Permission Ambiguities
- Who can perform this action?
- What authorization rules?
- Example: "Can unconfirmed users log in or should we block them?"
6. Performance Ambiguities
- What are acceptable response times?
- Pagination requirements?
- Example: "How many login attempts allowed per minute?"
7. Error Handling Ambiguities
- What errors can occur?
- How should errors be displayed?
- Example: "What message for duplicate email: '409 Conflict' or custom message?"
8. State Management Ambiguities
- How is data persisted?
- What about caching?
- Example: "Should we remember failed login attempts across sessions?"
9. Edge Case Ambiguities
- What about empty data?
- Race conditions?
- Example: "What if two users register with same email simultaneously?"
10. Testability Ambiguities (NEW)
- Are acceptance criteria specific enough to test?
- What edge cases need test coverage?
- What test data/fixtures are required?
- What should unit vs integration tests cover?
- Example: "Should we test with unicode characters in email addresses?"
Optional: Clarify Routing Preference
If routing preference is unclear or missing from spec, ask:
- "Is this feature simple, medium, complex, or expert-level?"
- "Should it use free/cheap/good/genius models?"
- "Any specific AI model requirements?"
Record the answer and update spec.md metadata.
Step 3: Present Questions
For each ambiguity found (max 5 most critical):
## Question N: [Topic]
**Context:** [Quote relevant spec section]
**What we need to know:** [Specific question]
**Testability Impact:** [How this affects test design]
**Options:**
A. [Option 1] - [Implications + how to test]
B. [Option 2] - [Implications + how to test]
C. [Option 3] - [Implications + how to test]
D. Custom - [Your answer]
**Your choice:** _Step 4: Update Specification
After receiving answers:
- Update spec.md with clarifications
- Remove
[NEEDS CLARIFICATION]markers - Add concrete details based on answers
- Ensure updated requirements are testable (Given/When/Then)
Step 5: Save Clarification Log
Create relentless/features/NNN-feature/clarification-log.md:
# Clarification Log: Feature Name
## Q1: [Topic] - RESOLVED
**Date:** 2026-01-11
**Question:** [Question]
**Answer:** [User's choice]
**Updated Sections:** [List spec sections updated]
**Test Impact:** [What tests can now be written]
## Q2: [Topic] - DEFERRED
**Date:** 2026-01-11
**Question:** [Question]
**Reason:** Can be decided during implementationExample
## Question 1: Password Requirements
**Context:** Spec says "password must be secure" but doesn't define requirements.
**What we need to know:** What are the specific password requirements?
**Testability Impact:** Need exact rules to write validation tests.
**Options:**
A. Basic (min 8 characters) - Simple, user-friendly. Test: reject 7 chars, accept 8.
B. Moderate (min 8 chars + number + symbol) - Balanced security. Test: reject "password", accept "Pass1!"
C. Strict (min 12 chars + number + symbol + upper/lower) - High security. Test: comprehensive regex.
D. Custom - Define your own rules
**Your choice:** B
---
## Question 2: Failed Login Handling
**Context:** Spec doesn't mention what happens after failed login attempts.
**What we need to know:** How should we handle repeated failed logins?
**Testability Impact:** Need exact limits to write rate limiting tests.
**Options:**
A. No limit - Allow unlimited attempts. Test: N/A (no limit to test)
B. Rate limit - Max 5 attempts per minute per IP. Test: 6th attempt in 60s fails.
C. Account lockout - Lock account after 5 failed attempts for 30 minutes. Test: verify lockout and unlock timing.
D. Custom - Define your own approach
**Your choice:** CNotes
- Prioritize by impact: security > scope > UX > technical
- Maximum 5 questions per session
- Some ambiguities can be deferred to implementation
- Update spec immediately after clarification
- Keep clarification log for future reference
- Every resolved ambiguity should enable writing tests
Other versions82 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 82
- antonio-orionus/Arroxy★ 610same name
- spool-lab/spool★ 592same name
- paperboytm/spool★ 590same name
- talesofai/cohub★ 571same name
- ZaxbyHub/opencode-swarm★ 517same name
- zaxbysauce/opencode-swarm★ 503same name
- nexmoe/DayBee★ 424same name
- nexmoe/eve★ 421same name
- educlopez/ui-craft★ 398same name
- ryanthedev/code-foundations★ 381same name
72 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
Used alongsideProjects with it also use quieter, delight and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
HK-hubSkillquieter
Tones down visually aggressive or overstimulating designs, reducing intensity while preserving quality. Use when the user mentions too bold, too loud, overwhelming, aggressive, garish, or wants a calmer, more refined aesthetic.
Also in 28 projects that use this
HK-hubSkilldelight
Add moments of joy, personality, and unexpected touches that make interfaces memorable and enjoyable to use. Elevates functional to delightful. Use when the user asks to add polish, personality, animations, micro-interactions, delight, or make an interface feel fun or memorable.
Also in 27 projects that use this
zwy123zwySkillnormalize
将用户随意输入的提示词规范化为结构清晰、便于大模型执行的标准化提示词。适用于需要整理、统一提示词格式时调用。
Also in 18 projects that use this
owshq-mecSkilllayout
Improve layout, spacing, and visual rhythm. Fixes monotonous grids, inconsistent spacing, and weak visual hierarchy. Use when the user mentions layout feeling off, spacing issues, visual hierarchy, crowded UI, alignment problems, or wanting better composition.
Also in 8 projects that use this
jovijoviSkillextract
Extract content from specific URLs using Tavily's extraction API. Returns clean markdown/text from web pages. Use when you have specific URLs and need their content without writing code.
Also in 19 projects that use this
- +7 more from HK-hub
In this repository15 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- analyzeSkill
- checklistSkill
- constitutionSkill
- convertSkill
- implementSkill
- learnSkill
- planSkill
- prdSkill
- relentlessSkill
- specifySkill
- tasksSkill
- taskstoissuesSkill
- validatorsSkill
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
Reviews can't be read right now.
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.

