eval-harness

eval-harness is a skill for Claude Code from sangrokjung/claude-forge. It costs 19 tokens per session (1,340 once invoked), scanned A, a copy of eval-harness, MIT.

A formal evaluation framework for coding-agent sessions based on eval-driven development, where expected behavior is defined and checked before or during implementation. It supports capability checks, regression checks, graders, and pass@k reliability measurements.

In plain words
What is it for?
Use it to define success criteria, run code-based or other evaluations, measure results, and track regressions across coding-agent changes.
Why use it?
It turns subjective judgments about an agent's work into repeatable checks. This helps show whether a new ability works and whether later changes break existing behavior.

Skill for Claude Code

Written for Claude Code: allowed-tools in frontmatter. Also seen: mentions Claude Code.

Part of the claude-forge plugin — 33 skills, 39 commands, 17 agents shipped together

Good fit Use it to define success criteria, run code-based or other evaluations, measure results, and track regressions across coding-agent changes.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/sangrokjung/claude-forge/eval-harness
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add sangrokjung/claude-forge --skill eval-harness
Clone the repo
git clone --depth 1 https://github.com/sangrokjung/claude-forge

Made for: Claude Code.

Or install claude-forge, the plugin that ships this one along with the rest of its 33 skills, 39 commands, 17 agents.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for eval-harness

README.md
[![agentmods](https://agentmods.dev/badge/skills/sangrokjung/claude-forge/eval-harness/github.svg)](https://agentmods.dev/skills/sangrokjung/claude-forge/eval-harness)
Your own site
<a href="https://agentmods.dev/skills/sangrokjung/claude-forge/eval-harness"><img src="https://agentmods.dev/badge/skills/sangrokjung/claude-forge/eval-harness/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for eval-harness

Your own site · 80×15
<a href="https://agentmods.dev/skills/sangrokjung/claude-forge/eval-harness"><img src="https://agentmods.dev/badge/skills/sangrokjung/claude-forge/eval-harness.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 19 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,340 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin 86% copy Near-identical to another mod in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00019 $0.01340
Opus 5 $0.00010 $0.00670
Sonnet 5 $0.00004 $0.00268
Haiku 4.5 $0.00002 $0.00134

Measured 5d ago against content hash 36fea82497c4, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

eval-harness scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

Origin

This is a copy

86% identical to eval-harness — 10 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.

skills/eval-harness/SKILL.md · 228 lines

How it starts

The opening of the file, as written. The whole thing — 228 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Eval Harness Skill

A formal evaluation framework for Claude Code sessions, implementing eval-driven development (EDD) principles.

Philosophy

Eval-Driven Development treats evals as the "unit tests of AI development":

  • Define expected behavior BEFORE implementation
  • Run evals continuously during development
  • Track regressions with each change
  • Use pass@k metrics for reliability measurement

Eval Types

Capability Evals

Test if Claude can do something it couldn't before:

[CAPABILITY EVAL: feature-name]
Task: Description of what Claude should accomplish
Success Criteria:
  - [ ] Criterion 1
  - [ ] Criterion 2
  - [ ] Criterion 3
Expected Output: Description of expected result

Regression Evals

Ensure changes don't break existing functionality:

[REGRESSION EVAL: feature-name]
Baseline: SHA or checkpoint name
Tests:
  - existing-test-1: PASS/FAIL
  - existing-test-2: PASS/FAIL
  - existing-test-3: PASS/FAIL
Result: X/Y passed (previously Y/Y)

Grader Types

1. Code-Based Grader

Deterministic checks using code:

# Check if file contains expected pattern
grep -q "export function handleAuth" src/auth.ts && echo "PASS" || echo "FAIL"

# Check if tests pass
npm test -- --testPathPattern="auth" && echo "PASS" || echo "FAIL"

# Check if build succeeds
npm run build && echo "PASS" || echo "FAIL"

2. Model-Based Grader

Use Claude to evaluate open-ended outputs:

[MODEL GRADER PROMPT]
Evaluate the following code change:
1. Does it solve the stated problem?
2. Is it well-structured?
3. Are edge cases handled?
4. Is error handling appropriate?

Score: 1-5 (1=poor, 5=excellent)
Reasoning: [explanation]

3. Human Grader

Flag for manual review:

[HUMAN REVIEW REQUIRED]
Change: Description of what changed
Reason: Why human review is needed
Risk Level: LOW/MEDIUM/HIGH

Metrics

pass@k

"At least one success in k attempts"

  • pass@1: First attempt success rate
  • pass@3: Success within 3 attempts
  • Typical target: pass@3 > 90%

Read the full file on GitHub · 228 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 5d ago First seen · 228 lines · 19 tokens per session scan A 36fea82497c4

Subscribe to this mod's changes

eval-harness is a skill published in the GitHub repository sangrokjung/claude-forge (825 stars, last pushed 5d ago), licensed MIT. It adds 19 tokens to every session and 1,340 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. It is 86% identical to eval-harness, differing in 10 lines, and is treated as a copy.

Related

Other skills, from other repositories

unit-test

A Go testing workflow for writing unit tests: small tests that check individual functions or components. It supports table-driven cases, where many inputs and expected results are organised in one test, and subtests.

johnqtcg/awesome-skills · 100 tokens

api-integration-test

Create, maintain, and run gated Go integration tests for internal APIs and service-to-service clients (HTTP/gRPC). Use for endpoint verification, contract checks with real runtime config, opt-in execution, timeout/retry safety, and integration failure triage in Go services.

johnqtcg/awesome-skills · 58 tokens

load-test

Performance load testing specialist for writing k6/vegeta/wrk scripts, defining SLOs, modeling scenarios (spike/soak/stress/breakpoint), analyzing results, and identifying bottlenecks. ALWAYS use when writing load test scripts, reviewing test results, designing test scenarios, setting performance SLOs, or diagnosing…

johnqtcg/awesome-skills · 121 tokens

e2e-test

Design, maintain, and execute reliable end-to-end tests for critical user journeys with Agent Browser as first choice for exploration and Playwright as the preferred code path for suites and CI. Use for E2E strategy, journey coverage, flaky test triage, artifact collection, CI gating, regression prevention, and…

johnqtcg/awesome-skills · 70 tokens

fuzzing-test

A Go testing guide for generating fuzz tests, which repeatedly try varied inputs to find crashes and unexpected behavior. It first checks whether the code is suitable for fuzzing.

johnqtcg/awesome-skills · 74 tokens

tdd-workflow

Enforce practical Test-Driven Development for code changes in Go services. Use for new features, bug fixes, refactors, API changes, and new modules. Requires Red-Green-Refactor evidence, defect-hypothesis-driven tests, killer cases, and coverage gates (line + risk-path).

johnqtcg/awesome-skills · 65 tokens