Code Grader — Testing & QA agent for Claude Code
Independent read-only reviewer that grades finished SDD work against its approved SPEC.md and the project's conventions.
How to install Code Grader
Installs to ~/.claude/agents/illescasdaniel-agent-guild-code-grader.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/illescasDaniel/agent-guild/HEAD/agents/code-grader.md -o ~/.claude/agents/illescasdaniel-agent-guild-code-grader.md Restart Claude Code, or start a new session, for it to be picked up.
What Code Grader does
name: code-grader description: Independent read-only reviewer that grades finished SDD work against its approved SPEC.md and the project's conventions. Use after Phase 4 implementation is green, before reporting the feature done. Pass the spec path and diff base. model: opus effort: high disallowedTools: Edit, Write, NotebookEdit, mcp__jevmem__memory_write, mcp__jevmem__memory_forget, mcp__jevmem__memory_pin, mcp__jevmem__memory_resolve, mcp__jevmem__memory_dismiss, mcp__jevmem__memory_conso
Alternatives in Testing & QA
- Ad Onboard Reviewer — Independent reviewer for AutoDeploy model onboarding 14.5k ★
- Fable Verifier — Cold verifier for finished fable deliverables 837 ★
- Integration Test Generator — Generates one integration/E2E test from an approved test case spec using existing project patterns 548 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Fpm QA Grader
Grades transcript quality. Checks coverage, confidence distribution, gaps, speaker consistency. Produces A-F g
Claude Grader
Grades one extension test run against its assertions and writes grading.json. Use after an eval run finishes,
Harmony Spec Implementation
Use this agent (normally spawned by harmony-goal in SDD Phase 4) to execute the implementation tasks of an app
Ccd Speckit Specifier
Use this agent to translate approved user intent into a testable, implementation-independent feature specifica
Story QA
Verify-phase crew member that runs a story's eval set against the DEV story through ./scripts/storyline eval-r
Playwright Test Reviewer
Read-only auditor that checks a given file set against this repository's test conventions. Use after any test
Related Skills
Refute
Get an independent reviewer to try to refute a claim that work is done
ExecuteTask
Phase 6 — implement tasks in order, choosing the loop (core / red-green-refactor / prompt-iteration) per task;
Assemble
Assemble the film — narrate, ASR-verify pronunciation, build, test, check — looping until green or reporting i