vc:stress-test

Pre-implementation risk surface. Two modes — pick by what you need:

• --predict (or default for "is this design sound?") — 5 expert personas debate the change
• --scenario (for "what could break?") — 12-dimension edge case decomposition

Mode: --predict (5-persona debate)

Five expert personas independently analyze a proposed change, then debate conflicts to produce a consensus verdict before a single line of code is written.

When to Use

• Before implementing a major or high-risk feature
• Before a significant refactor or architecture change
• Evaluating competing technical approaches
• Stress-testing assumptions in a proposed design

When NOT to Use

• Trivial or low-risk changes (use debugger for bugs, generate-plan for already-decided tasks)
• Already-approved work with no open design questions

The 5 Personas

Debate Protocol

1. Read the proposed change/feature description from the argument
2. Read relevant code if file paths are provided
3. Each persona analyzes independently
4. Identify agreements (all 4-5 align)
5. Identify conflicts and weigh tradeoffs
6. Produce verdict: GO | CAUTION | STOP with actionable recommendations

Output Format

## Prediction Report: [proposal title]

## Verdict: GO | CAUTION | STOP

### Agreements (all personas align)
- [Point 1]
- [Point 2]

### Conflicts & Resolutions

| Topic | Architect | Security | Performance | UX | Devil's Advocate | Resolution |
|-------|-----------|----------|-------------|-----|-----------------|------------|
| [Issue] | [View] | [View] | [View] | [View] | [View] | [Recommendation] |

### Risk Summary

| Risk | Severity | Mitigation |
|------|----------|------------|
| [Risk] | Critical/High/Medium/Low | [Action] |

### Recommendations
1. [Action item]
2. [Action item]
3. [Action item]

STOP Triggers (any one is sufficient)

• Security persona identifies auth bypass or data exposure with no viable mitigation
• Architect identifies fundamental design incompatibility requiring significant rework
• Performance persona identifies unacceptable latency or query explosion with no workaround
• Devil's Advocate exposes a false assumption that invalidates the entire approach

Mode: --scenario (12-dimension edge case decomposition)

Decompose any feature or code path across 12 dimensions to surface edge cases, risks, and test targets before implementation begins.

When to Use

• Before implementing complex or stateful features
• Before writing tests (generates test targets)
• Risk assessment during planning or code review
• API design review — surface contract edge cases early

When NOT to Use

• Trivial single-line changes or cosmetic UI tweaks
• Already well-tested, stable code with no recent modifications
• Pure configuration changes with no logic paths

The 12 Decomposition Dimensions

Workflow

1. Read target file(s) or parse feature description from argument
2. Filter dimensions — mark which of the 12 apply; skip irrelevant ones explicitly
3. Generate 3-5 scenarios per relevant dimension
4. Categorize severity — Critical / High / Medium / Low
5. Output as structured table

Severity Criteria

• Critical — data loss, security breach, auth bypass, silent corruption
• High — feature broken for a subset of users, data inconsistency
• Medium — degraded UX, recoverable error not surfaced to user
• Low — minor visual glitch, non-blocking warning

Output Format

## Scenario Report: [target]

Dimensions analyzed: [list]
Dimensions skipped: [list + reason]

| # | Dimension | Scenario | Severity | Expected Behavior |
|---|-----------|----------|----------|-------------------|
| 1 | Input Extremes | Empty string for required name field | High | Return 400 with field error |
| 2 | Authorization | Expired JWT accessing protected route | Critical | Redirect to login, invalidate session |
| 3 | Timing | Two users submit same form simultaneously | High | Idempotency key or conflict error |

### Summary
- Critical: N
- High: N
- Medium: N
- Low: N
- Total: N scenarios across X dimensions

When to use --predict vs --scenario

You can chain them: run --scenario first to find the top risks, then run --predict on the top risks as the change proposal.

Integration with Other Skills

End every response with the subagent status block:

**Status:** DONE | DONE_WITH_CONCERNS | BLOCKED | NEEDS_CONTEXT
**Summary:** [1-2 sentence summary]
**Mode run:** --predict | --scenario
**Verdict:** GO | CAUTION | STOP (for --predict) or scenario count by severity (for --scenario)
**Concerns/Blockers:** [if applicable]