Claude Comparator — Project Management agent for Claude Code
Blind-compares two outputs for the same task (labeled A and B) and says which is better and why.
How to install Claude Comparator
Installs to ~/.claude/agents/syncwin-ai-extender-claude-comparator.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/syncwin/ai-extender/HEAD/agents/claude-comparator.md -o ~/.claude/agents/syncwin-ai-extender-claude-comparator.md Restart Claude Code, or start a new session, for it to be picked up.
What Claude Comparator does
name: claude-comparator description: Blind-compares two outputs for the same task (labeled A and B) and says which is better and why. Use to compare an extension version against a baseline or a previous version without revealing which is which. model: inherit color: blue tools: ["Read", "Grep", "Glob"] metadata: displayName: "AI Extender Comparator" version: 1.0.0 author: "@wasimness" company: SyncWin plugin: ai-extender
You are the Claude implementation of the AI Extender com
Alternatives in Project Management
- Super Dev — Activate the Super Dev pipeline for research-first, commercial-grade project delivery 273 ★
- Agent Protocol & Architectural Mandate — Version: 2.10.5 Target Project: git-mcp-server Last Updated: 2026-03-25 This document defines the operational 206 ★
- Grainulator — Autonomous research sprint subagent 87 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
GitHub Project Sync
Syncs feature documentation with GitHub Issues. Analyzes docs/business-context/features/, compares against exi
Effort Grader
Blind grader for effortmining. Grades one artifact against a task prompt and a fixed rubric ONLY. The input pa
Team Extent Verifier
Geometry-regression verifier in a directing-agent-teams project. Compares every part of a new build against th
Decision Scorer
Scores 2+ architectural alternatives against PROJECT.md criteria. Called by architect after proposing variants
Power Verifier
POWER independent verifier. Judges one task's diff and check report against the plan, without the author's sum
Task Preparer
Reviews and clarifies tasks in Tasks/Next.md for better actionability
Related Skills
Load Coo Context
by Mjvolk3 - References specific files for sparse matrix operations, explains transform usage, compares with p
Prompt Test
Test LLM prompts against sample inputs. Shows outputs, checks for regressions when prompts change, and compare
Contextses
Показать контекст предыдущих сессий. Use when user says "что обсуждали", "продолжи с прошлого раза", "контекст