parallel-builder
Divide-and-conquer implementation from specs/plans. Decomposes a reference document into independent tasks, assigns each to a builder agent, executes in parallel waves respecting dependencies, then integrates results. Use when you have a spec, PRD, plan, or large feature to implement quickly with parallel execution.
What this skill does
# Parallel Builder: Divide-and-Conquer Implementation
Decomposes a plan/spec into independent tasks, spawns builder agents to implement each piece in parallel, and integrates the results. Unlike feature-council (same task, diverse approaches), this skill gives each agent a **different piece** of the work for maximum speed.
**Use this when you have a spec, PRD, or plan and want fast parallel implementation.**
## When to Use
โ
**Use parallel-builder for:**
- Implementing a PRD or spec document
- Large features that can be decomposed into parts
- When speed matters more than approach diversity
- Building from a reference file or detailed plan
- Multi-file implementations with clear boundaries
โ **Don't use for:**
- Bugs (use debug-council)
- Features where you want diverse approaches (use feature-council)
- Tasks that can't be parallelized
- Simple single-file changes
---
## Where Parallel-Builder Shines
### ๐ Maximum Speedup: Multi-File Specs
Parallel-builder is **fastest** when your spec produces **multiple independent files**:
```
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ IDEAL: Each agent owns different files โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ โ
โ Wave 2: TRUE PARALLEL EXECUTION โ
โ โโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโ
โ โ Agent 1 โ โ Agent 2 โ โ Agent 3 โโ
โ โ userService.ts โ โ authService.ts โ โ apiClient.ts โโ
โ โโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโ
โ โ โ โ โ
โ [PARALLEL - All 3 run simultaneously] โ
โ โ
โ Speedup: ~3ร faster than sequential โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
```
**Great use cases:**
- Full-stack features (types + services + routes + UI components)
- Microservice implementations (multiple independent services)
- CRUD APIs (each resource in its own files)
- Plugin/module systems (each module independent)
- Test suites (test files for different modules)
### โ ๏ธ Falls Back to Sequential: Single-File Specs
When multiple tasks modify the **same file**, parallel execution would cause conflicts. The skill automatically falls back to sequential:
```
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ LIMITED: All agents modify same file โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ โ
โ Wave 3: SEQUENTIAL (conflict avoidance) โ
โ โโโโโโโโโโโโโโโโโโโ โ
โ โ Agent 1 โ โ modifies BG600L.cpp โ
โ โโโโโโโโโโฌโโโโโโโโโ โ
โ โ โ
โ โโโโโโโโโโโโโโโโโโโ โ
โ โ Agent 2 โ โ modifies BG600L.cpp โ
โ โโโโโโโโโโฌโโโโโโโโโ โ
โ โ โ
โ โโโโโโโโโโโโโโโโโโโ โ
โ โ Agent 3 โ โ modifies BG600L.cpp โ
โ โโโโโโโโโโโโโโโโโโโ โ
โ โ
โ No speedup, but still provides organized task breakdown โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
```
**Still useful for:**
- Organized task breakdown and tracking
- Clear separation of concerns
- Systematic implementation of complex single-file changes
### Speedup Estimate
| Spec Type | Parallel Potential | Expected Speedup |
|-----------|-------------------|------------------|
| Multi-file, independent modules | High | 50-80% faster |
| Multi-file, some shared | Medium | 30-50% faster |
| Few files, many functions | Low | 10-20% faster |
| Single file, many changes | Minimal | ~0% (sequential) |
The skill will show you the estimated speedup in the execution plan before you confirm.
---
## Step 0: Accept Plan Input
First, get the plan from the user. Accept one of:
1. **File path**: User provides path to a spec, PRD, or plan file
2. **Inline description**: User describes what to build
3. **Reference file**: User points to existing code to replicate/extend
```
What would you like me to build? You can provide:
- A file path to a spec/PRD/plan
- A description of the feature
- A reference file to work from
I'll decompose it into parallel tasks for fast implementation.
```
---
## Step 1: Analyze & Decompose
Read the plan and identify work units. For each unit, determine:
1. **Independence**: Can it be built without waiting for others?
2. **Dependencies**: What must exist before this can be built?
3. **Outputs**: What files/interfaces does this create?
4. **Complexity**: Rough estimate (small/medium/large)
### Decomposition Rules
| Rule | Description |
|------|-------------|
| **File boundaries** | One agent per file or closely related file group |
| **Feature boundaries** | One agent per distinct feature/capability |
| **Layer boundaries** | Separate by layer (types, services, API, UI) |
| **Minimize coupling** | Tasks should have minimal interface dependencies |
### Create Dependency Graph
Identify three categories:
```
WAVE 1 (Foundation - No Dependencies):
- Shared types/interfaces
- Configuration
- Constants
- Base utilities
WAVE 2 (Core - Depends on Wave 1):
- Services
- Business logic
- Data access
- Core components
WAVE 3+ (Integration - Depends on Previous):
- API routes that use services
- UI that uses components
- Integration code
- Tests
```
---
## Step 2: Define Shared Contracts
**CRITICAL**: Before spawning agents, define the contracts they share.
Create a shared context that ALL agents receive:
```markdown
## Shared Contracts
### Interfaces
[Define all shared types/interfaces that multiple agents need]
### File Ownership
[Map which agent owns which files - no overlaps]
### Naming Conventions
[Consistent naming across all agents' work]
### Import Paths
[How agents should import from each other's domains]
```
This prevents conflicts and ensures pieces fit together.
---
## Step 3: Present Execution Plan to User
Show the decomposition and get confirmation:
```
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
PARALLEL BUILD PLAN
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
๐ Source: [file path or "inline description"]
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
TASK BREAKDOWN
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
## Wave 1: Foundation (3 agents in parallel)
| Task | Agent | Files | Complexity |
|------|-------|-------|------------|
| Create shared types | builder-solver-1 | types/*.ts | Small |
| Database schema | builder-solver-2 | schema.prisma | Medium |
| Config & constants | builder-solver-3 | config/*.ts | Small |
## Wave 2: Core (2 agents in parallel, after Wave 1)
| Task | Agent | Files | Depends On |
|------|-------|-------|------------|
| User service | builder-solver-4 | services/user.ts | Wave 1 types |
| Auth service | builder-solver-5 | services/auth.ts | Wave 1 types |
## Wave 3: Integration (1 agent, after Wave 2)
| Task | Agent | Files | Depends On |
|------|-------|-------|------------|
| API routes + wiring | builder-solver-6 | routes/*.ts, index.ts | Wave 1, 2 |
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
SUMMARY
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
Total Tasks: 6
Total Waves: 3
Parallel Agents: Up to 3 concurrent
Estimated Speedup: ~50-60% faster than sequential
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโRelated in AI Agents
skill-development
IncludedComprehensive meta-skill for creating, managing, validating, auditing, and distributing Claude Code skills and slash commands (unified in v2.1.3+). Provides skill templates, creation workflows, validation patterns, audit checklists, naming conventions, YAML frontmatter guidance, progressive disclosure examples, and best practices lookup. Use when creating new skills, validating existing skills, auditing skill quality, understanding skill architecture, needing skill templates, learning about YAML frontmatter requirements, progressive disclosure patterns, tool restrictions (allowed-tools), skill composition, skill naming conventions, troubleshooting skill activation issues, creating custom slash commands, configuring command frontmatter, using command arguments ($ARGUMENTS, $1, $2), bash execution in commands, file references in commands, command namespacing, plugin commands, MCP slash commands, Skill tool configuration, or deciding between skills vs slash commands. Delegates to docs-management skill for official documentation.
reprompter
IncludedTransform messy prompts into well-structured, effective prompts โ single or multi-agent. Use when: "reprompt", "reprompt this", "clean up this prompt", "structure my prompt", rough text needing XML tags and best practices, "reprompter teams", "repromptception", "run with quality", "smart run", "smart agents", multi-agent tasks, audits, parallel work, anything going to agent teams. Don't use when: simple Q&A, pure chat, immediate execution-only tasks. See "Don't Use When" section for details. Outputs: Structured XML/Markdown prompt, quality score (before/after), optional team brief + per-agent sub-prompts, agent team output files. Success criteria: Single mode quality score โฅ 7/10; Repromptception per-agent prompt quality score 8+/10; all required sections present, actionable and specific.
adaptive-compaction
IncludedAdaptive add-on policy and recovery layer that decides WHEN to compact, prune, snapshot, or fork -- replacing fixed-percent auto-compaction across Claude Code, Codex, and MCP-capable hosts. Trigger on auto-compact timing or damage: "when should I compact", "is it safe to compact now or start a fresh session", "auto-compact fires too early/mid-task", "switching to an unrelated task but the window still has space", "context rot", "answers get worse the longer the session runs", "the agent forgot the plan or my decisions after it summarized", "add a layer on top that manages context without changing the agent", raising autoCompactWindow to give the policy room, or installing/tuning a cross-tool compaction policy or PreCompact hook -- even when "compaction" is never said but the problem is context-window pressure or post-summarization memory loss. Do NOT use to summarize a conversation, build RAG, write a summarization prompt (decides WHEN not HOW), or answer max-context-length trivia.
agent-skill-creator
IncludedCreate cross-platform agent skills from workflow descriptions. Activates when users ask to create an agent, automate a repetitive workflow, create a custom skill, or need advanced agent creation. Triggers on phrases like create agent for, automate workflow, create skill for, every day I have to, daily I need to, turn process into agent, need to automate, create a cross-platform skill, validate this skill, export this skill, migrate this skill. Supports single skills, multi-agent suites, transcript processing, template-based creation, interactive configuration, cross-platform export, and spec validation.
llm-wiki
IncludedUse when building or maintaining a persistent personal knowledge base (second brain) in Obsidian where an LLM incrementally ingests sources, updates entity/concept pages, maintains cross-references, and keeps a synthesis current. Triggers include "second brain", "Obsidian wiki", "personal knowledge management", "ingest this paper/article/book", "build a research wiki", "compound knowledge", "Memex", or whenever the user wants knowledge to accumulate across sessions instead of being re-derived by RAG on every query.
skill-master
IncludedAgent Skills authoring, evaluation, and optimization. Create, edit, validate, benchmark, and improve skills following the agentskills.io specification. Use when designing SKILL.md files, structuring skill folders (references, scripts, assets), ingesting external documentation into skills, running trigger evals, benchmarking skill quality, optimizing descriptions, or performing blind A/B comparisons. Keywords: agentskills.io, SKILL.md, skill authoring, eval, benchmark, trigger optimization.