pre-task-diagnostic

Pre-Task Diagnostic Skill

Analyze a task before execution to surface hidden risks and dependencies.

When to Use

• Before delegating tasks to a sub-agent
• Before starting complex feature implementation
• When task scope is unclear
• Before assigning work in parallel

Input Types

• Feature description
• Issue ID
• User story
• Bug report

Diagnostic Process

1. Scope Analysis

• What is the task trying to achieve?
• What are the boundaries?
• Is the scope well-defined or ambiguous?

2. Dependency Detection

Search codebase for:

• Related files that might need changes
• Services/modules that will be affected
• API contracts that might change
• Database migrations needed

3. Risk Assessment

Evaluate:

• Breaking change potential
• Security implications
• Performance impact
• Test coverage gaps

4. Blocker Identification

Check for:

• Missing requirements
• Pending PRs that block this work
• External dependencies (API, design, etc.)
• Technical debt that needs addressing first

5. Agent Routing Recommendation

Based on analysis, recommend:

• Dialog (Claude #1): High complexity, needs discussion
• Parallel (Claude #2): Medium complexity, can run independently
• Async agent: Low complexity, well-defined, independent

Output Format

# Pre-Task Diagnostic: [Task Name]

## Task Summary
**Input:** [Original task description]
**Interpreted Scope:** [What this task actually involves]

## Scope Assessment

| Aspect | Status | Notes |
|--------|--------|-------|
| Clarity | OK/WARN/FAIL | [Is the task well-defined?] |
| Boundaries | OK/WARN/FAIL | [Are boundaries clear?] |
| Acceptance Criteria | OK/WARN/FAIL | [Are success criteria defined?] |

**Scope Size:** Small / Medium / Large
**Estimated Complexity:** Low / Medium / High

## Dependencies

### Files Likely Affected
| Layer | Files | Impact |
|-------|-------|--------|
| Domain | [files] | [changes needed] |
| Infrastructure | [files] | [changes needed] |
| UseCase | [files] | [changes needed] |
| Handler/API | [files] | [changes needed] |
| Presentation | [files] | [changes needed] |

### External Dependencies
- [ ] [API changes needed?]
- [ ] [Design assets needed?]
- [ ] [Other team coordination?]

### Blocking Issues
- [ ] [PR #xxx must merge first]
- [ ] [Issue #xxx must be resolved]

## Risk Assessment

| Risk | Likelihood | Impact | Mitigation |
|------|------------|--------|------------|
| [Risk 1] | High/Med/Low | High/Med/Low | [How to mitigate] |

### Breaking Changes
- [ ] API contract changes
- [ ] Database schema changes
- [ ] Interface changes

### Security Considerations
- [Any security implications?]

## Go/No-Go Decision

| Criterion | Status |
|-----------|--------|
| Scope is clear | OK/FAIL |
| Dependencies identified | OK/FAIL |
| No blockers | OK/FAIL |
| Risks are acceptable | OK/FAIL |

**Recommendation:** GO / NO-GO / NEEDS CLARIFICATION

## Agent Routing

**Recommended Agent:** Dialog / Parallel / Async

**Reasoning:** [Why this agent is best suited]

## Pre-Flight Checklist

Before starting:
- [ ] Run `implementation-sequencer` for optimal implementation ordering
- [ ] Confirm acceptance criteria with stakeholder
- [ ] Create Decision Artifact in `.claude/decisions/`

## Related Tasks

If this task is too large, consider decomposing:
- [Subtask 1]
- [Subtask 2]

Guidelines

• Use Grep/Glob to search for related code
• Check git history for recent changes in affected areas
• Flag scope creep risks explicitly
• Be conservative with "GO" decisions

Related Skills

• implementation-sequencer: For optimal implementation ordering
• flow-parallel: When routing to parallel implementation
• pre-task-diagnostic: Recursively for subtasks

Gotchas

See gotchas.md in this directory for known pitfalls and recurring mistakes when using this skill.