Agent skill

sanity-check

Step back, reflect on current work, validate direction and alignment. Use when complexity is increasing, feeling uncertain, before major decisions, or something feels off.

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/data/sanity-check

SKILL.md

/sanity-check

Mid-work validation using reflection to catch drift early. Permission to pause and think.

Usage

bash
/sanity-check                        # Reflect on current work
/sanity-check --project coordinatr   # Focus on specific project

When to Use

  • Complexity is increasing
  • Feeling uncertain about direction
  • Before major decisions
  • Something feels off
  • After 30+ minutes of planning work

Not for: Session start (use /refresh), after completion (just continue)

Execution Steps

1. Reflect Using Sequential Thinking

Process these questions:

  • What are we trying to accomplish?
  • What have we done so far?
  • What's the current approach?
  • Does this align with the project vision?
  • Are we solving the right problem?
  • What concerns exist?

Categorize findings:

  • Green: On track, continue
  • Yellow: Minor issues, easy fixes
  • Red: Major drift, course correction needed

2. Read Context Files

bash
# Project context (if specified)
Read: ideas/{project}/README.md
Read: ideas/{project}/project-brief.md
Read: ideas/{project}/critique.md

# General context
Read: CLAUDE.md
Read: about-me.md

# Recent work
Bash: git log -5 --format="%h - %s"

Skip missing files gracefully.

3. Analyze Alignment

Check against:

  • Vision: Does current work support project goals?
  • Brief: Are we addressing the stated problem?
  • Critique: Are we avoiding known pitfalls?
  • Patterns: Are we following repo conventions?

4. Provide Assessment

markdown
## Sanity Check

### Current State
[What we're working on, current approach]

### Alignment Check
- **Vision**: [status] [brief assessment]
- **Problem Fit**: [status] [brief assessment]
- **Approach**: [status] [brief assessment]

### Concerns

**What's Working**
- [Positive observation]

**Minor Issues**
- [Yellow flag + suggested fix]

**Critical Issues**
- [Red flag + required action]

### Recommendation
[Continue as-is | Minor adjustment | Course correction | Pause and discuss]

### Next Steps
1. [Specific action]
2. [Specific action]

Philosophy

  • Permission to pause: Makes stepping back a legitimate workflow step
  • Catch drift early: Course correction cheap now, expensive later
  • Trust your gut: If something feels off, run this command

Didn't find tool you were looking for?

Be as detailed as possible for better results