Eval Grader
Independent skill-eval grader with no review authorship and no fixture access. Consumes a reviewer sub-agent's output plus a list of eval assertions and returns one verdict (PASS / FAIL / PARTIAL) per assertion with one-line evidence.
- Type
- Subagent
- Repository
- serpro69/claude-toolbox
- GitHub stars
- 149
- License
- Not declared
- Repo last updated
- Sep 25, 2026
- Source file
- klaude-plugin/agents/eval-grader.md
- Model
- claude-opus-4-8[1m]
What Eval Grader is
Eval Grader is a subagent published in the serpro69/claude-toolbox repository on GitHub, which has about 149 stars. The repository describes itself as: “Minimal by design, explicitly multi-lang, production-ready and battle-tested collection of configs and plugins for your Claude Code agentic development workflows: mcp, configs, skills, agents and more.”
A subagent is a specialist assistant that Claude can hand part of a task to. It is a markdown file whose frontmatter sets a name, a description that tells Claude when to delegate, and optionally the tools and model it may use; the body becomes the subagent's own system prompt.
Because a subagent works in its own context, it keeps the main conversation focused: Claude can send a narrow job, such as a review or a specialised analysis, to Eval Grader and get back a compact result.
How to install Eval Grader
Claude Code
- Download eval-grader.md from the repository.
- Save it to ~/.claude/agents/ to use it in every project, or to .claude/agents/ inside one project to share it through version control.
- Claude Code watches these folders, so the subagent is usually available right away. Ask Claude to use it by name, or @-mention it to make sure it runs.
Claude Cowork
- Cowork loads subagents through plugins. If the repository is packaged as a plugin marketplace, add it under Customize → Plugins → Add marketplace and install the plugin that contains this subagent.
- Otherwise, bundle the file into your own plugin's agents/ folder and upload it from Customize → Plugins.
New to extending Cowork? Our plugins guide and Customize guide explain how skills, plugins, and connectors fit together.
Inside the source file
The repository does not declare an open-source license, so we list only the outline of the file here. Read the full text on GitHub.
- What You Receive
- What You Do NOT Have
- Tools Policy
- Mandatory ordering — exempt
- How To Grade
- Output Format
- What To Avoid
Before you install
- Read the whole file first. Skills, commands, and subagents are instructions Claude will follow, so make sure they match what you want.
- Check which tools, scripts, or MCP servers it uses. Local servers and scripts run with your permissions.
- Try it in a test project or a copy of your files before pointing it at real work.
- Pin the version you tested, and review changes before updating.
- Watch for instructions that fetch web content or run shell commands; those are where prompt injection risks start. See our prompt injection guide.
FAQ
What is Eval Grader?
Eval Grader is a subagent for Claude Code and Claude Cowork from the serpro69/claude-toolbox repository on GitHub. Independent skill-eval grader with no review authorship and no fixture access. Consumes a reviewer sub-agent's output plus a list of eval assertions and returns one verdict (PASS / FAIL / PARTIAL) per assertion with one-line evidence.
How do I install Eval Grader in Claude Code?
Download eval-grader.md from the repository. Save it to ~/.claude/agents/ to use it in every project, or to .claude/agents/ inside one project to share it through version control. Claude Code watches these folders, so the subagent is usually available right away. Ask Claude to use it by name, or @-mention it to make sure it runs.
Can I use Eval Grader in Claude Cowork?
Cowork loads subagents through plugins. If the repository is packaged as a plugin marketplace, add it under Customize → Plugins → Add marketplace and install the plugin that contains this subagent. Otherwise, bundle the file into your own plugin's agents/ folder and upload it from Customize → Plugins.
Is Eval Grader safe to install?
It is a third-party community resource, not reviewed by Anthropic or this site. Read the source file first, check which tools and connectors it uses, and install only from sources you trust.
Similar resources
- Design Reviewer Independent design document reviewer with no authorship attachment. Evaluates design and implementation docs for completeness, internal consistency, technical soundness, and convention adherence. Subagent · serpro69/claude-toolbox
- Kk Development workflow skills, commands, and hooks from claude-toolbox Plugin · serpro69/claude-toolbox
- Spec Reviewer Independent spec conformance reviewer with no authorship attachment. Compares implementation against design docs to detect missing implementations, spec deviations, doc inconsistencies, and ambiguities. Subagent · serpro69/claude-toolbox
- Override Agent Agent with a model override in the manifest. Subagent · serpro69/claude-toolbox
- Profile Resolver Resolves active profiles and their checklist-load decisions for a given diff, using the shared profile-detection procedure. Emits a structured resolution report: active profiles with per-file `triggered_by` signal, plus loaded vs. not-loaded checklists with Load-if reasoning. Useful wherever profile detection must be isolated from downstream work — the /kk:review-code eval harness uses it to… Subagent · serpro69/claude-toolbox
- Code Reviewer Independent code reviewer with no authorship attachment. Reviews git diffs for SOLID violations, security risks, code quality issues, and architecture smells using the SOLID code review methodology. Subagent · serpro69/claude-toolbox
- Pneuma Impl Fable Pneuma Skills implementation specialist — HEAVYWEIGHT Fable-5 engine. Identical discipline to pneuma-impl (TDD + contract-first… Subagent · pandazki/pneuma-skills
- Acceptance Criteria Agent Expert in drafting clear, testable acceptance criteria for requirements and user stories, ensuring traceability to business goals. Use for comprehensive acceptance criteria development. Subagent · ChrisRoyse/610ClaudeSubagents