AgentSkillsCN

context-aware-questions

主动识别项目背景中的信息缺口,并提出切实可行的问题,以挖掘遗漏的需求、模糊不清的规格,或亟待填补的文档空白。

SKILL.md
--- frontmatter
name: context-aware-questions
description: "Proactively identify information gaps in project context and generate actionable questions to surface missing requirements, unclear specifications, or documentation gaps"
tags:
  - questions
  - gaps
  - requirements
  - analysis
  - documentation

Context-Aware Question Generator Skill

Analyze project context to surface information gaps and generate prioritized, actionable questions.

When to Use

Activate when:

  • User asks "what am I missing?" or "what questions should I be asking?"
  • Reviewing issues before starting work
  • Analyzing documentation completeness
  • Preparing a draft (issue, PR, document) for submission
  • During project planning or kickoff

Available Tools

This skill leverages the following tools:

generate_project_questions

Scans the entire project context (issues, notes, project scan) for information gaps.

Best for: Holistic project review, "what's missing overall?"

generate_issue_questions

Analyzes a specific GitHub issue for missing requirements or unclear specifications.

Best for: Pre-work review, issue clarification

suggest_clarifications

Reviews a draft (issue, PR, note) before submission and suggests improvements.

Best for: Quality check before posting

identify_information_gaps

Scans a specific target (README, documentation, codebase) for gaps.

Best for: Targeted documentation review

Output Structure

Questions are categorized and prioritized:

Categories

  • requirements: Missing acceptance criteria, unclear specs, vague descriptions
  • technical: Implementation questions, architecture decisions
  • documentation: Missing docs, unclear README sections
  • process: Workflow gaps, missing labels/assignees
  • consistency: Conflicting information across sources
  • architecture: High-level design questions (requires project scan)

Priorities

  • high: Blocking or critical gaps that need immediate attention
  • medium: Important gaps that should be addressed soon
  • low: Nice-to-have clarifications

Example Workflows

1. Project-Wide Review

code
User: "What questions should I be asking about this project?"

Assistant uses: generate_project_questions

Output:
## Information Gaps Found

### High Priority (3)
1. [requirements] Issue #42 lacks acceptance criteria
   - Context: "Add login feature" has no definition of done
   - Suggested Action: Add checkboxes with acceptance criteria

2. [documentation] No API documentation found
   - Context: Project has 15 endpoints but no API reference
   - Suggested Action: Create docs/api.md with endpoint documentation

3. [requirements] Issue #38 has vague description
   - Context: Description is only 20 characters
   - Suggested Action: Expand with problem statement and expected behavior

### Medium Priority (5)
...

2. Issue Review Before Starting Work

code
User: "Are there any gaps in issue #42?"

Assistant uses: generate_issue_questions with issue_number: 42

Output:
## Issue #42 Analysis

### Questions to Clarify

1. [requirements] What are the acceptance criteria?
   - Issue lacks checkboxes or "done when" statement

2. [technical] What authentication method should be used?
   - No mention of OAuth, JWT, or session-based auth

3. [process] Who should be assigned to this issue?
   - Issue has no assignee

3. Draft Review Before Submission

code
User: "Help me improve this issue before I submit it:
Title: Bug in login
Body: Login doesn't work"

Assistant uses: suggest_clarifications with draft_type: 'issue'

Output:
## Suggested Clarifications

1. [requirements] What are the acceptance criteria?
   - Add an "Acceptance Criteria" section with checkboxes

2. [requirements] Can you provide more detail?
   - The description is very brief (18 chars)
   - Add: problem description, expected behavior, context

3. [requirements] How can this bug be reproduced?
   - Add "Steps to Reproduce" section

Improved draft:
---
## Problem
[Describe what's happening]

## Expected Behavior
[What should happen instead]

## Steps to Reproduce
1. Go to /login
2. Enter credentials
3. Click submit
4. Observe error

## Acceptance Criteria
- [ ] Login works with valid credentials
- [ ] Appropriate error shown for invalid credentials

4. Documentation Gap Analysis

code
User: "What's missing from our documentation?"

Assistant uses: identify_information_gaps with target: 'documentation'

Output:
## Documentation Gaps

1. [documentation] No docs/ directory found
   - Suggested Action: Create docs/ with API reference, guides

2. [documentation] README lacks usage examples
   - No code blocks or example snippets found

3. [documentation] No contribution guidelines
   - No CONTRIBUTING.md or contributing section in README

Gap Detection Rules

Issue Analysis

  • Missing acceptance criteria (no checkboxes, no "done when")
  • Vague descriptions (< 100 chars, contains "unclear", "might", "maybe")
  • Missing labels or assignees
  • Stale issues (open > 30 days with no activity)
  • Bug reports without reproduction steps

Documentation Analysis

  • Missing README sections (install, usage, examples)
  • No dedicated docs/ directory
  • Short README (< 500 chars)
  • Missing license information
  • No inline documentation (JSDoc/docstrings)

Project-Wide Analysis

  • Inconsistencies between issues and notes
  • Missing architecture documentation
  • Undocumented key files
  • No project scan available

Best Practices

When Generating Questions

  1. Always prioritize by impact
  2. Group related questions together
  3. Provide actionable suggested actions
  4. Link to specific sources (issue number, file path)

When Reviewing Drafts

  1. Be constructive, not critical
  2. Suggest specific improvements
  3. Provide example text when helpful
  4. Acknowledge what's already good

When Analyzing Documentation

  1. Check against industry standards
  2. Consider the project type (CLI needs different docs than library)
  3. Leverage project scan data when available