Essentials
Essential tooling for discovering docs, cleaning up code, deep thinking (ultraplan), and systematic code review.
Features:
#### UltraPlan - Deep Investigation Methodology
Systematic multi-step investigation of complex problems with confidence-based progression.
```bash
/ultraplan I need help designing a better architecture for...
```
What it does:
- Progressive confidence tracking (exploring β low β medium β high β certain)
- Multi-step investigation with state accumulation
- Iterative Haiku agent guidance for next-step suggestions
- Final Sonnet agent validation for comprehensive analysis
- Practical engineering focus with trade-off analysis
- Implementation options with pros/cons (similar to Claude Code plan mode)
#### CodeReview - Systematic Code Analysis
Multi-step code review with severity-classified findings and actionable fixes.
```bash
/codereview src/auth.py src/middleware.py
```
What it does:
- Progressive confidence tracking (exploring β low β medium β high β certain)
- Severity classification for all issues (π΄ CRITICAL, π HIGH, π‘ MEDIUM, π’ LOW)
- Multi-step investigation with state accumulation
- Iterative Haiku agent guidance for coverage validation
- Final Sonnet agent validation for independent review
- Actionable fixes with before/after code examples
- Top 3 priorities with effort estimates
- Quick wins vs. long-term improvements
Focus areas:
- Security vulnerabilities (SQL injection, XSS, hardcoded secrets, auth gaps)
- Concurrency issues (race conditions, deadlocks, thread-safety)
- Resource management (memory leaks, unclosed connections)
- Error handling gaps
- Performance bottlenecks (algorithmic complexity, N+1 queries)
- Code quality and maintainability
Best for:
- Security audits before production
- Bug investigation and root cause analysis
- Performance bottleneck identification
- Code quality assessment
- Pre-deployment validation
#### Documentation Lookup Skill
Fetches library documentation via Context7's HTTP API using the "code execution" pattern.
What it does:
- Proactively activates when you mention libraries or frameworks
- Writes curl commands to fetch documentation from Context7 API
- Two-step process: search for library ID, then fetch docs
- Falls back to web search if needed
- Use: "Check the docs for..."
Why code execution? Instead of always loading MCP tool definitions into context, Claude writes API calls only when neededβreducing context usage and demonstrating the pattern from [Anthropic's code execution blog post](https://www.anthropic.com/engineering/code-execution-with-mcp).
#### Code Examples Skill
Find real-world code examples across millions of GitHub repositories using [grep.app](https://grep.app).
What it does:
- Finds how others implement specific patterns in production code
- Filters by language, repository, or file path
- Use: "How do others implement X?" or "Show me usage examples for..."
Complements documentation-lookup: While documentation-lookup fetches official docs, code-examples finds real-world usage in production code.
#### Code Janitor Agent
Mechanical code maintenance specialist for tedious cleanup work.
What it does:
- Fixes linting errors systematically
- Renames variables consistently across files
- Updates import statements
- Fixes formatting issues
- Iterates until all fixable issues are resolved
#### Consensus - Multi-Perspective Decision Analysis
Analyze complex decisions from multiple viewpoints using specialized debate agents.
```bash
/consensus Should we migrate to microservices?
```
What it does:
- Launches three parallel Sonnet agents with distinct analytical stances:
- FOR - Advocates for the proposal with optimistic perspective
- AGAINST - Critiques the proposal rigorously
- NEUTRAL - Provides objective analysis weighing evidence
- Presents balanced view of trade-offs and considerations
- Helps avoid confirmation bias in decision-making
|