# okf-comparison-reader

The okf-comparison-reader subagent for Claude Code, Codex, and Cursor checks task answers against references. It evaluates queries, rules, numbers, and answers stating that the bundle contains no information.

- subagent · `type=agent` · Claude Code, Codex, Cursor · ★ 105 repo stars
- Author: wmahfoudh
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/wmahfoudh/okf/tree/HEAD/skills-pi/okf-produce/agents/okf-comparison-reader.md
- Updated: 2026-10-31
- Security check: automated, 2026-11-17 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/agents/okf-comparison-reader

## What it does

okf-comparison-reader issues a verdict for each task by comparing its question, reference answer, and derived answer. It receives no full bundle or producer assessment. Its available tools are `read`, `grep`, `find`, `ls`, and `bash`.

For query tasks, it compares result rows in light of tables, joins, filters, and parameters. If the database can be reached, it runs both queries and compares their output. Otherwise, it reviews the query text and says the result needs manual confirmation.

## Install

### Claude Code (native platform)

**Manually: put the files in place**
- File: `.claude/agents/okf-comparison-reader.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/okf-comparison-reader.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/okf-comparison-reader.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/wmahfoudh/okf/tree/HEAD/skills-pi/okf-produce/agents/okf-comparison-reader.md

## When to use it
- compare each task answer with its reference
- check whether two queries return matching rows
- judge answers about rules and numbers

## Example prompts
- Compare the task answers with their references and give a verdict for each one.
- Check whether the two queries in this task return the same rows.
- Judge this answer about a rule against its reference.

## Similar
- [log-reader](https://affpapa.org/en/nevibecode/skills/log-reader.md) — skill · 145 installs
- [gaia-architecture-comparison](https://affpapa.org/en/nevibecode/skills/gaia-architecture-comparison.md) — skill · 433 installs
- [hyperliquid-reader](https://affpapa.org/en/nevibecode/skills/hyperliquid-reader.md) — skill · 598 installs
- [android-benchmark-comparison](https://affpapa.org/en/nevibecode/skills/android-benchmark-comparison.md) — skill · 559 installs
- [esg-benchmark-comparison](https://affpapa.org/en/nevibecode/skills/esg-benchmark-comparison.md) — skill · 276 installs
- [okf-knowledge-base](https://affpapa.org/en/nevibecode/skills/okf-knowledge-base.md) — skill · 9400 installs

From the same set: [wmahfoudh/okf](https://affpapa.org/en/nevibecode/by/wmahfoudh/okf.md) — 8 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
