# analyzer

analyzer examines blind skill comparisons to explain the winner’s advantage and suggest changes to the losing skill. Run it after the comparator finishes.

- subagent · `type=agent` · Claude Code, Codex, Cursor · ★ 406 repo stars
- Author: syahiidkamil
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/syahiidkamil/Software-Engineer-AI-Agent-Atlas/tree/HEAD/.claude/skills/skill-creator/agents/analyzer.md
- Updated: 2026-06-25
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/agents/analyzer

## What it does

The subagent examines the comparator’s result alongside both skills and their execution transcripts. It looks for differences that could explain the outcome, including how clearly each skill guides the agent, which tools the agent uses, and how it handles errors. It needs the comparison result and both transcripts to carry out this analysis.

The output is a JSON report with a comparison summary, winner strengths, loser weaknesses, instruction-following scores, and proposed changes. Suggestions can cover instructions, tools, examples, error handling, structure, or references, with a priority assigned to each. This gives you a concrete revision plan for the losing skill.

## Install

### Claude Code (native platform)

**Manually: put the files in place**
- File: `.claude/agents/analyzer.md` — For all projects use ~/.claude/agents/. Claude delegates by the description field or an @-mention

### Codex

**Manually: put the files in place**
- File: `.codex/agents/analyzer.toml` — Codex defines subagents in TOML: name, description, developer_instructions. Move the Markdown text into developer_instructions by hand. For all projects use ~/.codex/agents/

### Cursor

**Manually: put the files in place**
- File: `.cursor/agents/analyzer.md` — Cursor also reads .claude/agents/

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/syahiidkamil/Software-Engineer-AI-Agent-Atlas/tree/HEAD/.claude/skills/skill-creator/agents/analyzer.md

## When to use it
- explain why one skill won a blind comparison
- spot gaps between skill instructions and agent actions
- score how well each agent followed its skill
- prioritize revisions to the losing skill

## Example prompts
- Compare the two skill runs and save your analysis to `analysis.json`.
- Explain why the skill at `winner_skill_path` won, then suggest revisions for the other skill.
- Review both execution transcripts and score instruction following.

## Similar
- [perf-analyzer](https://affpapa.org/en/nevibecode/agents/perf-analyzer.md) — subagent · ★ 2003 repo stars
- [gc-log-analyzer](https://affpapa.org/en/nevibecode/skills/gc-log-analyzer.md) — skill · 1 install
- [historical-cost-analyzer](https://affpapa.org/en/nevibecode/skills/historical-cost-analyzer.md) — skill · 118 installs
- [yao-meta-skill](https://affpapa.org/en/nevibecode/skills/yao-meta-skill.md) — skill · 686 installs
- [evaluate-skill](https://affpapa.org/en/nevibecode/skills/edonadei-evaluate-skill.md) — skill · ★ 237 repo stars
- [skill-doctor](https://affpapa.org/en/nevibecode/skills/skill-doctor.md) — skill · 15 installs

From the same set: [syahiidkamil/software-engineer-ai-agent-atlas](https://affpapa.org/en/nevibecode/by/syahiidkamil/software-engineer-ai-agent-atlas.md) — 26 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
