agentic-eval

github/awesome-copilot · updated Apr 10, 2026

MDX-style export adds YAML metadata + attribution linking explainx.ai and this canonical listing URL.

$npx skills add https://github.com/github/awesome-copilot --skill agentic-eval
0 commentsdiscussion
summary

Iterative evaluation and refinement patterns for improving AI agent outputs through self-critique loops.

  • Provides three core patterns: basic reflection (self-critique loops), evaluator-optimizer (separated generation and evaluation), and code-specific test-driven refinement
  • Supports multiple evaluation strategies including outcome-based assessment, LLM-as-judge comparison, and rubric-based scoring with weighted dimensions
  • Includes practical Python implementations with structured JSON
skill.md

Agentic Evaluation Patterns

Patterns for self-improvement through iterative evaluation and refinement.

Overview

Evaluation patterns enable agents to assess and improve their own outputs, moving beyond single-shot generation to iterative refinement loops.

Generate → Evaluate → Critique → Refine → Output
    ↑                              │
    └──────────────────────────────┘

When to Use

  • Quality-critical generation: Code, reports, analysis requiring high accuracy
  • Tasks with clear evaluation criteria: Defined success metrics exist
  • Content requiring specific standards: Style guides, compliance, formatting

Pattern 1: Basic Reflection

Agent evaluates and improves its own output through self-critique.

def reflect_and_refine(task: str, criteria: list[str], max_iterations: int = 3) -> str:
    """Generate with reflection loop."""
    output = llm(f"Complete this task:\n{task}")
    
    for i in range(max_iterations):
        # Self-critique
        critique = llm(f"""
        Evaluate this output against criteria: {criteria}
        Output: {output}
        Rate each: PASS/FAIL with feedback as JSON.
        """)
        
        critique_data = json.loads(critique)
        all_pass = all(c["status"] == "PASS" for c in critique_data.values())
        if all_pass:
            return output
        
        # Refine based on critique
        failed = {k: v["feedback"] for k, v in critique_data.items() if v["status"] == "FAIL"}
        output = llm(f"Improve to address: {failed}\nOriginal: {output}")
    
    return output

Key insight: Use structured JSON output for reliable parsing of critique results.


Pattern 2: Evaluator-Optimizer

Separate generation and evaluation into distinct components for clearer responsibilities.

class EvaluatorOptimizer:
    def __init__(self, score_threshold: float = 0.8):
        self.score_threshold = score_threshold
    
    def generate(self, task: str) -> str:
        return llm(f"Complete: {task}")
    
    def evaluate(self, output: str, task: str) -> dict:
        return json.loads(llm(f"""
        Evaluate output for task: {task}
        Output: {output}
        Return JSON: {{"overall_score": 0-1, "dimensions": {{"accuracy": ..., "clarity": ...}}}}
        """))
    
    def optimize(self, output: str, feedback: dict) -> str:
        return llm(f"Improve based on feedback: {feedback}\nOutput: {output}")
    
    def run(self, task: str, max_iterations: int = 3) -> str:
        output = self.generate(task)
        for _ in range(max_iterations):
            evaluation = self.evaluate(output, task)
            if evaluation["overall_score"] >= self.score_threshold:
                break
            output = self.optimize(output, evaluation)
        return output

Pattern 3: Code-Specific Reflection

Test-driven refinement loop for code generation.

class CodeReflector:
    def reflect_and_fix(self, spec: str, max_iterations: int = 3) -> str:
        code = llm(f"Write Python code for: {spec}")
        tests = llm(f"Generate pytest tests for: {spec}\nCode: {code}")
        
        for _ in range(max_iterations):
            result = run_tests(code, tests)
            if result["success"]:
                return code
            code = llm(f"Fix error: {result['error']}\nCode: {code}")
        return code

Evaluation Strategies

Outcome-Based

Evaluate whether output achieves the expected result.

def evaluate_outcome(task: str, output: str, expected: str) -> str:
    return llm(f"Does output achieve expected outcome? Task: {task}, Expected: {expected}, Output: {output}")

LLM-as-Judge

Use LLM to compare and rank outputs.

def llm_judge(output_a: str, output_b: str, criteria: str) -> str:
    return llm(f"Compare outputs A and B for {criteria}. Which is better and why?")

Rubric-Based

Score outputs against weighted dimensions.

RUBRIC = {
    "accuracy": {"weight": 0.4},
    "clarity": {"weight": 0.3},
    "completeness": {"weight": 0.3}
}

def evaluate_with_rubric(output: str, rubric: dict) -> float:
    scores = json.loads(llm(f"Rate 1-5 for each dimension: {list(rubric.keys())
how to use agentic-eval

How to use agentic-eval on Cursor

AI-first code editor with Composer

1

Prerequisites

Before installing skills in Cursor, ensure your development environment meets these requirements:

  • Cursor installed and configured on your development machine
  • Node.js version 16.0+ with npm package manager (verify with node --version)
  • Active project directory or workspace where you want to add agentic-eval
2

Execute installation command

Execute the skills CLI command in your project's root directory to begin installation:

$npx skills add https://github.com/github/awesome-copilot --skill agentic-eval

The skills CLI fetches agentic-eval from GitHub repository github/awesome-copilot and configures it for Cursor.

3

Select Cursor when prompted

The CLI will show a list of available agents. Use arrow keys to navigate and space to select Cursor:

◆ Which agents do you want to install to?
│ ── Universal (.agents/skills) ── always included ────
│ • Amp
│ • Antigravity
│ • Cline
│ • Codex
│ ●Cursor(selected)
│ • Cursor
│ • Windsurf
4

Verify installation

Confirm successful installation by checking the skill directory location:

.cursor/skills/agentic-eval

Reload or restart Cursor to activate agentic-eval. Access the skill through slash commands (e.g., /agentic-eval) or your agent's skill management interface.

Security & Verification Notice

We perform automated surface-level scans (Gen AI Scanner, Socket, Snyk) during installation. These checks detect common vulnerabilities but do not guarantee complete security. Always review skill source code and verify the publisher's reputation before production use.

Skills execute code in your development environment. Always verify the publisher's identity, review recent commits, and test in isolated environments before production deployment.

List & Monetize Your Skill

Submit your Claude Code skill and start earning

GET_STARTED →

Use Cases

User Story & Requirements Generation

Create detailed user stories, acceptance criteria, and feature specs

Example

Generate user stories for 'password reset feature' with acceptance criteria, edge cases, and test scenarios

Reduce spec writing time by 50%, ensure comprehensive coverage

Competitive Analysis

Research competitors, compare features, identify gaps

Example

Analyze 5 competitor products, create feature comparison matrix, suggest differentiation opportunities

Complete competitive research in 2 hours instead of 2 days

Roadmap Prioritization

Evaluate features using frameworks (RICE, ICE, Kano) and create prioritized backlogs

Example

Score 20 feature ideas using RICE framework, generate prioritized roadmap with rationale

Make data-driven prioritization decisions faster

Stakeholder Communication

Draft PRDs, status updates, and stakeholder presentations

Example

Create executive summary of Q3 roadmap, monthly progress report, feature launch announcement

Save 3-5 hours/week on communication overhead

Implementation Guide

Prerequisites

  • Claude Desktop or compatible AI client
  • Access to product documentation and roadmap tools (Jira, Notion, etc.)
  • Understanding of product management frameworks (RICE, Jobs-to-be-Done, etc.)
  • Stakeholder contact information and communication channels

Time Estimate

30-60 minutes to see productivity improvements

Installation Steps

  1. 1.Install product management skill
  2. 2.Start with user story generation for known feature
  3. 3.Progress to competitive analysis: research 2-3 competitors
  4. 4.Use for roadmap prioritization: apply RICE/ICE scoring
  5. 5.Draft stakeholder communications and refine based on feedback
  6. 6.Build template library for recurring PM tasks
  7. 7.Share effective prompts with product team

Common Pitfalls

  • Not validating competitive research—verify facts before sharing
  • Accepting user stories without involving engineering team
  • Over-relying on frameworks without qualitative judgment
  • Not customizing outputs to company culture and communication style
  • Skipping stakeholder validation of generated requirements

Best Practices

✓ Do

  • +Validate research and competitive analysis with real data
  • +Collaborate with engineering when generating technical requirements
  • +Customize frameworks and templates to your company context
  • +Use skill for first drafts, refine with stakeholder input
  • +Document successful prompt patterns for PM tasks
  • +Combine AI efficiency with human judgment and intuition

✗ Don't

  • Don't publish competitive analysis without fact-checking
  • Don't finalize user stories without engineering review
  • Don't make prioritization decisions solely on AI scoring
  • Don't skip customer validation of generated requirements
  • Don't ignore company-specific context and culture

💡 Pro Tips

  • Provide context: company goals, constraints, customer feedback
  • Ask for alternatives: 'Show 3 ways to prioritize this roadmap'
  • Request stakeholder-specific formatting: 'Executive summary vs. engineering spec'
  • Use skill for 70% generation + 30% customization to company needs

When to Use This

✓ Use When

Use for user story writing, competitive research, roadmap prioritization, stakeholder communication, and PRD drafting. Best for reducing repetitive documentation and research work.

✗ Avoid When

Avoid for strategic product vision (requires deep customer empathy), pricing decisions (needs market and financial expertise), or when face-to-face customer discovery is more valuable than speed.

Learning Path

  1. 1Basic: user stories, feature specs, status updates
  2. 2Intermediate: competitive analysis, prioritization frameworks, PRDs
  3. 3Advanced: product strategy, go-to-market planning, OKR setting
  4. 4Expert: product vision, market positioning, business model innovation

Discussion

Product Hunt–style comments (not star reviews)
  • No comments yet — start the thread.
general reviews

Ratings

4.652 reviews
  • Xiao Liu· Dec 28, 2024

    agentic-eval fits our agent workflows well — practical, well scoped, and easy to wire into existing repos.

  • Kiara Gill· Dec 20, 2024

    Registry listing for agentic-eval matched our evaluation — installs cleanly and behaves as described in the markdown.

  • Li Nasser· Dec 16, 2024

    Keeps context tight: agentic-eval is the kind of skill you can hand to a new teammate without a long onboarding doc.

  • Amelia Thompson· Nov 19, 2024

    agentic-eval fits our agent workflows well — practical, well scoped, and easy to wire into existing repos.

  • Kiara Ghosh· Nov 15, 2024

    agentic-eval is among the better-maintained entries we tried; worth keeping pinned for repeat workflows.

  • Zaid Iyer· Nov 11, 2024

    agentic-eval reduced setup friction for our internal harness; good balance of opinion and flexibility.

  • Meera Chawla· Nov 7, 2024

    I recommend agentic-eval for anyone iterating fast on agent tooling; clear intent and a small, reviewable surface area.

  • Hassan Taylor· Oct 26, 2024

    agentic-eval reduced setup friction for our internal harness; good balance of opinion and flexibility.

  • Chen Khanna· Oct 10, 2024

    agentic-eval is among the better-maintained entries we tried; worth keeping pinned for repeat workflows.

  • Noah Torres· Oct 6, 2024

    agentic-eval fits our agent workflows well — practical, well scoped, and easy to wire into existing repos.

showing 1-10 of 52

1 / 6