/plan-with-team
Source: ~/.claude/skills/plan-with-team/SKILL.md
name: plan-with-team description: Create implementation plans with builder/validator agent teams. Use for major features, refactoring, or system components. argument-hint: "[what to build]"
Plan With Team
Create an implementation plan with builder/validator agent teams.
Related Skills (pick the right one)
/plan-with-team(this skill) — Multi-expert planning for complex features, refactors, system components./build-plan— Execute the plan produced here. Run AFTER plan is approved./hop-build <task_id>— For a SINGLE task (not a multi-task plan)./build— Toggle session into Build Mode./prime-build— Load lightweight build context into session.
Instructions
You are creating a detailed implementation plan that assigns work to specialized agent teams.
Step 1: Research (MANDATORY — with Context Isolation)
Before planning, delegate research to specialist agents using ISOLATION PATTERN:
-
If building a company/org: Dispatch proxima with isolation preamble:
- "Research 2-3 existing examples, write FULL analysis to /tmp/{task_id}-research-deliverables.md"
- "Return ONLY: summary + deliverable path + memory_candidates"
-
If building software: Dispatch petter-graff/martin-kleppmann with isolation:
- "Explore codebase patterns, write FULL architecture analysis to deliverable"
- "Return structured summary only"
-
If building process: Dispatch sentinel-ba with isolation:
- "Research industry standards, write FULL report to deliverable"
Load isolation template: ~/system/prompts/SUBAGENT_ISOLATION.md
Collect summaries, Read deliverables ONLY for critical details needed in plan.
Step 2: Analyze
Based on research:
- What patterns should we copy?
- What should we adapt for our needs?
- What are the key components?
Step 3: Define Team
For each major task, assign:
- Builder — Creates/implements
- Validator — Verifies the work
Reference agents from ~/.claude/agents/:
builder.md— Implementation agentvalidator.md— Verification agent (read-only)
Step 4: Create Plan
Write plan to ~/system/specs/<name>-plan.md with this structure:
# Plan: [Name]
## Research Summary
[What we learned from existing examples]
## Objective
[1-2 sentences]
## Team Orchestration
### Team Members
| ID | Name | Role | Agent Type |
|----|------|------|------------|
| B1 | [name]-builder | Build [what] | builder |
| V1 | [name]-validator | Validate [what] | validator |
### Step-by-Step Tasks
#### Phase 1: [Name]
**Task 1:** [Description]
- Owner: B1
- BlockedBy: none
- Acceptance:
- [ ] Criterion 1
- [ ] Criterion 2
**Task 2:** Validate [above]
- Owner: V1
- BlockedBy: 1
- Acceptance: [criteria]
## Validation Commands
[How to verify the work]
Step 5: Validate Plan
Self-validate the plan:
- All tasks have acceptance criteria
- Dependencies make sense (validators blocked by builders)
- No circular dependencies
MANDATORY CHECK — plan is INCOMPLETE without both:
- Validation task exists — owner: Proveo/Angie Jones, end-to-end test with real evidence (not dry-run), BlockedBy all build tasks
- Documentation task exists — owner: Skillforge, BookStack page for every system built or changed, BlockedBy validation task
If either is missing → add them before presenting to CEO. Do not ask. Just add.
Step 5.5: Auto-create MC IDs for all tasks (Phase 4 enforcement)
After self-validating the plan, run the plan parser to create MC IDs for every Task block:
node ~/system/tools/plan-parser.js ~/system/specs/<plan-name>-plan.md --parent <master_mc_id>
This ensures Tasks 11+ (validation/docs) get first-class MC IDs visible in mc.js list.
Without this, validation/docs tasks become invisible and structurally susceptible to skip pattern.
After parser runs, verify:
- All Task N: blocks have MC: #
line under heading node ~/system/tools/mc.js list --owner john | grep "Plan task"shows new MC IDs
Step 6: Present for Approval
Show the user:
- Research summary
- Plan overview
- Number of tasks and phases
- Ask: "Approve plan? Then run
/build-planto execute."
Output
The plan file path: ~/system/specs/<name>-plan.md
Ready for execution with /build-plan.
$ARGUMENTS