AGENT SKILL

eval-harness

Formal evaluation framework for Claude Code sessions implementing eval-driven development (EDD) principles

License 许可
Apache-2.0
所属仓库
pjherron/hypoc
最近更新
2026-09-01
Stars
★ 0
Forks
0

01 前言 Front Matter

name
eval-harness
tools
Read, Write, Edit, Bash, Grep, Glob
metadata
{ "origin": "ECC" }
description
Formal evaluation framework for Claude Code sessions implementing eval-driven development (EDD) principles

03 SKILL.md 正文

Eval Harness Skill

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

When to Activate

  • Setting up eval-driven development (EDD) for AI-assisted workflows
  • Defining pass/fail criteria for Claude Code task completion
  • Measuring agent reliability with pass@k metrics
  • Creating regression test suites for prompt or agent changes
  • Benchmarking agent performance across model versions

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%

pass^k

"All k trials succeed"

  • Higher bar for reliability
  • pass^3: 3 consecutive successes
  • Use for critical paths

Eval Workflow

1. Define (Before Coding)

## EVAL DEFINITION: feature-xyz

### Capability Evals
1. Can create new user account
2. Can validate email format
3. Can hash password securely

### Regression Evals
1. Existing login still works
2. Session management unchanged
3. Logout flow intact

### Success Metrics
- pass@3 > 90% for capability evals
- pass^3 = 100% for regression evals

2. Implement

Write code to pass the defined evals.

3. Evaluate

# Run capability evals
[Run each capability eval, record PASS/FAIL]

# Run regression evals
npm test -- --testPathPattern="existing"

# Generate report

4. Report

EVAL REPORT: feature-xyz
========================

Capability Evals:
  create-user:     PASS (pass@1)
  validate-email:  PASS (pass@2)
  hash-password:   PASS (pass@1)
  Overall:         3/3 passed

Regression Evals:
  login-flow:      PASS
  session-mgmt:    PASS
  logout-flow:     PASS
  Overall:         3/3 passed

Metrics:
  pass@1: 67% (2/3)
  pass@3: 100% (3/3)

Status: READY FOR REVIEW

Integration Patterns

Pre-Implementation

/eval define feature-name

Creates eval definition file at .claude/evals/feature-name.md

During Implementation

/eval check feature-name

Runs current evals and reports status

Post-Implementation

/eval report feature-name

Generates full eval report

Eval Storage

Store evals in project:

.claude/
  evals/
    feature-xyz.md      # Eval definition
    feature-xyz.log     # Eval run history
    baseline.json       # Regression baselines

Best Practices

  1. Define evals BEFORE coding - Forces clear thinking about success criteria
  2. Run evals frequently - Catch regressions early
  3. Track pass@k over time - Monitor reliability trends
  4. Use code graders when possible - Deterministic > probabilistic
  5. Human review for security - Never fully automate security checks
  6. Keep evals fast - Slow evals don't get run
  7. Version evals with code - Evals are first-class artifacts

Example: Adding Authentication

## EVAL: add-authentication

### Phase 1: Define (10 min)
Capability Evals:
- [ ] User can register with email/password
- [ ] User can login with valid credentials
- [ ] Invalid credentials rejected with proper error
- [ ] Sessions persist across page reloads
- [ ] Logout clears session

Regression Evals:
- [ ] Public routes still accessible
- [ ] API responses unchanged
- [ ] Database schema compatible

### Phase 2: Implement (varies)
[Write code]

### Phase 3: Evaluate
Run: /eval check add-authentication

### Phase 4: Report
EVAL REPORT: add-authentication
==============================
Capability: 5/5 passed (pass@3: 100%)
Regression: 3/3 passed (pass^3: 100%)
Status: SHIP IT

Product Evals (v1.8)

Use product evals when behavior quality cannot be captured by unit tests alone.

Grader Types

  1. Code grader (deterministic assertions)
  2. Rule grader (regex/schema constraints)
  3. Model grader (LLM-as-judge rubric)
  4. Human grader (manual adjudication for ambiguous outputs)

pass@k Guidance

  • pass@1: direct reliability
  • pass@3: practical reliability under controlled retries
  • pass^3: stability test (all 3 runs must pass)

Recommended thresholds:

  • Capability evals: pass@3 >= 0.90
  • Regression evals: pass^3 = 1.00 for release-critical paths

Eval Anti-Patterns

  • Overfitting prompts to known eval examples
  • Measuring only happy-path outputs
  • Ignoring cost and latency drift while chasing pass rates
  • Allowing flaky graders in release gates

Minimal Eval Artifact Layout

  • .claude/evals/<feature>.md definition
  • .claude/evals/<feature>.log run history
  • docs/releases/<version>/eval-summary.md release snapshot

04 同仓库更多 Skills

pjherron
pjherron/hypoc

workspace-surface-audit

Audit the active repo, MCP servers, plugins, connectors, env surfaces, and harness setup, then recommend the highest-value ECC-native skills, hooks, agents, and operator workflows. Use when the user wants help setting up Claude Code or understanding what capabilities are actually available in their environment.

★ 0
pjherron
pjherron/hypoc

Verification & Quality Assurance

Comprehensive truth scoring, code quality verification, and automatic rollback system with 0.95 accuracy threshold for ensuring high-quality agent outputs and codebase reliability.

★ 0
pjherron
pjherron/hypoc

verification-loop

A comprehensive verification system for Claude Code sessions.

★ 0
pjherron
pjherron/hypoc

triage

Move issues and external PRs through a state machine of triage roles — categorise, verify, grill if needed, and write agent-ready briefs.

★ 0
pjherron
pjherron/hypoc

token-budget-advisor

Offers the user an informed choice about how much response depth to consume before answering. Use this skill when the user explicitly wants to control response length, depth, or token budget. TRIGGER when: "token budget", "token count", "token usage", "token limit", "response length", "answer depth", "short version", "brief answer", "detailed answer", "exhaustive answer", "respuesta corta vs larga", "cuántos tokens", "ahorrar tokens", "responde al 50%", "dame la versión corta", "quiero controlar cuánto usas", or clear variants where the user is explicitly asking to control answer size or depth. DO NOT TRIGGER when: user has already specified a level in the current session (maintain it), the request is clearly a one-word answer, or "token" refers to auth/session/payment tokens rather than response size.

★ 0
pjherron
pjherron/hypoc

to-tickets

Break a plan, spec, or the current conversation into a set of tracer-bullet tickets, each declaring its blocking edges, published to the configured tracker — edges as text in a local file, or native blocking links on a real tracker.

★ 0