From claude-setup
Interactive scoping, roadmap generation, and multi-agent validation
How this skill is triggered — by the user, by Claude, or both
Slash command
/claude-setup:roadmap-with-validationThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
Roadmap validation stress-tests the implementation plan before execution begins. Multiple validators review the roadmap looking for:
Roadmap validation stress-tests the implementation plan before execution begins. Multiple validators review the roadmap looking for:
Question: Does every V1 feature directly contribute to the success criteria?
For each feature:
Red Flags:
Question: Is the sequencing actually correct?
Tests:
Trace Test:
For each task:
Task X requires -> [list everything] -> verify each exists earlier
Red Flags:
Question: Are any tasks hiding significant complexity?
For each task:
Red Flags:
Question: Could someone verify each task without asking you?
For each criterion:
Red Flags:
Fix Pattern:
Before: "User can import recipes"
After: "User pastes URL -> system extracts title/ingredients/instructions ->
displays preview -> user confirms -> recipe appears in list within 3 seconds.
Error shown if URL invalid or parsing fails."
Categories:
For each risk:
If PARALLEL-READY:
Common Issues:
Questions:
# Roadmap Validation — [MODEL_NAME]
You are a skeptical technical reviewer. Your job is to stress-test this roadmap and find gaps before execution begins.
**Stance:** Assume something is wrong.
**Success Criteria:** [V1_SUCCESS_CRITERIA]
**Review the roadmap below and check:**
1. Success criteria alignment
2. Dependency validation
3. Complexity audit
4. Acceptance criteria quality
5. Risk assessment
6. Parallelization reality check (if applicable)
7. Out of scope validation
**Output Format:**
## Summary Verdict
[APPROVED / APPROVED WITH CHANGES / NEEDS REVISION]
## Key Findings
1. [Finding with severity: Critical/Important/Minor]
2. [Finding]
3. [Finding]
## Required Changes
- [ ] [Specific change needed]
- [ ] [Specific change needed]
## Risks Identified
| Risk | Severity | Mitigation | Phase |
|------|----------|------------|-------|
## Detailed Review
[Section-by-section analysis]
---
**ROADMAP TO REVIEW:**
[PASTE ROADMAP.md CONTENT]
Roadmap is solid. No blocking issues found.
Criteria:
Roadmap is workable but needs specific fixes.
Criteria:
Roadmap has fundamental issues requiring rework.
Criteria:
Critical (Must Fix):
Important (Should Fix):
Suggestions (Nice to Have):
Dependency Addition:
Before: Phase 1: User Profile
After: Phase 0: Foundation
- [ ] Authentication setup
Phase 1: User Profile (requires Auth)
Task Splitting:
Before:
- [ ] Build Dashboard
After:
- [ ] Build Dashboard Layout
- [ ] Add Dashboard Widgets
- [ ] Implement Dashboard Data Fetching
Acceptance Criteria Clarification:
Before:
- [ ] Implement Recipe Import
- Acceptance: User can import recipes
After:
- [ ] Implement Recipe Import
- Scope: URL paste, parsing, preview, confirmation
- Acceptance: User pastes URL -> system extracts title/ingredients/instructions
-> displays preview -> user confirms -> recipe appears in list
within 3 seconds
- Error Handling: Show error if URL invalid or parsing fails
- Verification: Manual test with 5 different recipe sites
After validation passes, add to ROADMAP.md:
> **Validated:** [DATE]
> **Validation Sources:** claude, gpt4, grok, gemini
> **Status:** APPROVED WITH CHANGES
> **Changes Applied:** [N] critical, [N] important
Skip validation if:
Don't skip if:
Protocol version: 1.0 | Created: 2026-02-01
npx claudepluginhub matthewod11-stack/claude-setup --plugin claude-setupCreates phased implementation plans with risks, dependencies, validation checkpoints, and rollback notes. Useful for non-trivial tasks involving multiple files, subsystems, strategies, or migrations.
Generates phased roadmaps (MVP, Scale, Advanced) from architecture designs with Epic/Story/Task breakdowns, T-shirt effort estimates, and validation gates.