Skip to main content
subagent CX Codex

grader — Claude Code subagent: what it does, how to install

grader assesses agent execution transcripts and generated files against expectations, flags superficial passes, and critiques test assertions.

no license ★ 406 updated 2026-06-25

Install into Claude Code · Manually: put the files in place All methods and platforms ↓
.claude/agents/grader.md
Open on GitHub

What grader does

The subagent audits task runs by comparing expected outcomes against an execution transcript and generated artifacts. It takes three parameters: a list of assertions, a path to the Markdown log, and an output directory. Instead of trusting what the executor claimed in the transcript, it opens and inspects output files directly. A test fails whenever an artifact exists with hollow or mismatched content, preventing fake passes.

Beyond prewritten checks, grader extracts unstated claims about process, facts, and output quality from the execution log and verifies them against the artifacts. If user_notes.md is present in the output folder, the agent reads executor uncertainties and notes them in the review. All ratings, tool call statistics, and timing benchmarks are saved into grading.json.

The burden of proof stays on the assertion. Absent clear evidence, the verdict defaults to FAIL. The subagent also reviews the eval definitions themselves, pointing out assertions that are too lenient to catch failures.

What it helps with

  • score task results against explicit expectations
  • catch superficial test passes with invalid contents
  • verify implicit factual and process claims from logs
  • pinpoint gaps and weak checks in evaluation suites
  • generate a structured grading.json report with timing metrics

Example prompts for the agent

  • Review the execution log in transcript.md and verify outputs in run_artifacts/
  • Grade this task run against my checklist and write the result to grading.json
  • Audit the latest session transcript and suggest improvements for weak eval assertions

How to install grader

Manually: put the files in place

.claude/agents/grader.md

For all projects use ~/.claude/agents/. Claude delegates by the description field or an @-mention

Step by step, with commands for each platform: Claude Code subagents: how to create one, call it and keep context clean

Questions about grader

How do I install grader in Claude Code?

Place the subagent definition into .claude/agents/grader.md within your project repository.

What does grader do?

It checks execution transcripts and generated files against expectations, assigns PASS or FAIL verdicts with evidence, and exports findings to grading.json.

How does grader handle uncertain or superficial results?

The subagent puts the burden of proof on the expectation; if evidence is weak or matches merely by coincidence, it marks the check as FAIL.

Similar subagents

All subagents for Testing and QA →

From the same set26

The whole set →
qa-manual-tester qa-manual-tester takes over a real browser and pokes at your app the way a tester would: clicks, typing, screenshots, console reads. Handy when a feature needs a hands-on pass. CX Codex ★ 406repo stars comparator comparator runs blind comparisons between two AI outputs against the original prompt. Identifies the stronger result without knowing which skill produced each file. CX Codex ★ 406repo stars code-review code-review — a subagent for Claude Code that reviews a pull request through gh and posts a comment listing confirmed bugs. Skips closed, draft, or already-reviewed PRs without prompting. CX Codex ★ 406repo stars change-core-self change-core-self asks you about the project, works out a fitting identity for the agent from scratch, and stores it as a rule file loaded on every session. CX Codex +1 ★ 406repo stars create-phase The create-phase command clarifies a development phase through Q&A, then creates an HTML wireframe with the decisions needed for implementation in Claude Code plan mode. CX Codex +1 ★ 406repo stars visual The visual command turns a discussion into a self-contained HTML document with diagrams, wireframes and matrices the agent implements from in plan mode. CX Codex +1 ★ 406repo stars oneshotapp oneshotapp builds a functioning app from a brief idea in Claude Code without human intervention. It logs architecture choices and verifies implementation slices on the fly. CX Codex +1 ★ 406repo stars 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. CX Codex ★ 406repo stars

Project sponsors

Companies that keep this analytics open