simplify-code

Reviews recent code changes with three parallel subagents and applies cleanup fixes.

Updated Jun 17, 2026
One-click install
npx skills add https://github.com/i-bebsi/hermes-agent --skill simplify-code-i-bebsi
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: simplify-code
Source: https://github.com/i-bebsi/hermes-agent/tree/main/hermes-config/skills/software-development/simplify-code
Command: npx skills add https://github.com/i-bebsi/hermes-agent --skill simplify-code-i-bebsi

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Reviewing your own recent changes for duplication, quality issues, and inefficiencies is slow and easy to do superficially. This Skill runs three focused reviewers in parallel over your git diff, aggregates their findings, and applies the fixes worth keeping. ## Core Features & Use Cases - Parallel Three-Agent Review: Launches reuse, quality, and efficiency reviewers concurrently via delegate_task batch mode, each searching the full codebase for evidence. - Aggregation and Conflict Resolution: Merges findings, discards false positives, resolves conflicts using a correctness-first priority order, and applies fixes with patch or write_file. - Flexible Scoping and Dry Run: Supports staged changes, last commit, branch diffs, specific files, focus filters (reuse, quality, efficiency), and report-only dry runs. - Use Case: After finishing a feature branch, say "simplify my changes" to get duplication, quality, and performance issues found and fixed, with targeted tests re-run to confirm nothing broke. ## Quick Start Ask the agent to simplify my recent changes and it will review the current git diff with three parallel reviewers and apply the worthwhile fixes.

Frequently Asked Questions about simplify-code

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I review and clean up my recent code changes automatically?▼

Invoke the skill with a phrase like "simplify my changes" and it captures your git diff, runs three parallel reviewers for reuse, quality, and efficiency, then applies the validated fixes and re-runs targeted tests.

How to run a code review without applying any changes?▼

Request a dry run by saying "simplify but don't change anything" or "just report". The three reviewers still run and present findings, but nothing is applied until you approve.

Can I scope the review to a specific commit or file?▼

Yes. You can target staged changes, the last commit, a branch diff against main, or specific files like src/foo.py. The skill selects the matching git diff command based on your request.

What happens when the parallel reviewers disagree on a fix?▼

Conflicts are resolved with a fixed priority: correctness first, then your stated focus, then readability and reuse, then micro-performance. Mutually exclusive suggestions default to the one touching less code.

What are the limitations of parallel multi-agent code review?▼

Very large diffs over roughly 2000 lines become token-heavy since each reviewer carries the full diff, so scoping down is recommended. The skill also avoids fan-out beyond three reviewers to limit cost and conflicting suggestions.