Skip to content
cccskills
FreePublished on GitHubEnglish overview

agentic-eval

Design workflows that evaluate and refine AI outputs through self-review, separate generation and evaluation roles, tests, or weighted scoring criteria. Includes patterns for structured feedback, iteration limits, and stopping when quality stops improving.

OriginalShow the original description

Patterns and techniques for evaluating and improving AI agent outputs. Use this skill when: - Implementing self-critique and reflection loops - Building evaluator-optimizer pipelines for quality-critical generation - Creating test-driven code refinement workflows - Designing rubric-based or LLM-as-judge evaluation systems - Adding iterative improvement to agent outputs (code, reports, analysis) - Measuring and improving agent response quality

LeoYeAI/agent-master-skills2,193 stars(GitHub 2,193 + cccskills 0)

Run score Not tested

See how to install

Good for

  • Adding self-review and revision loops to AI outputs
  • Separating generation from evaluation
  • Refining generated code using test results
  • Comparing responses against weighted scoring criteria

Overview

What it does

This skill helps design workflows that evaluate AI agent outputs and revise them based on the results. It presents patterns for self-critique, separating generation from evaluation, and using test results to refine code. It also covers checking outputs against expected outcomes, using an LLM as a judge to compare answers, and scoring with weighted criteria.

When it helps

Use it when code, reports, or analysis need review against defined standards after an initial draft. It is useful when success criteria such as accuracy, clarity, completeness, or formatting can be stated clearly, and you want to build an evaluation and improvement loop around them.

Example requests

  • “Design a loop that checks a report for accuracy and completeness, then fixes gaps.”
  • “Plan a workflow that improves generated code using pytest results.”
  • “Create a way to compare two answers with weighted scoring criteria.”

Notes

Examples using llm or run_tests need implementation for the environment where they will run. The source recommends setting an iteration limit, stopping when scores no longer improve, and recording evaluation history. It also uses JSON for evaluation results and recommends handling parse failures.

This overview was written by AI (OpenAI Codex) from the public SKILL.md. Check the original below for the exact behavior.

Document check

AI checked the SKILL.md for operations that need care and for setup requirements. The score comes only from actually running the skill (hands-on test).

SafetyAre sending data, deleting files, and running commands disclosed?
BPartly lackingA `run_tests` code-execution example is included, but runtime safeguards are not detailed.
Easy setupDoes it need few tools or API keys, and are they documented?
BPartly lackingEnvironment-specific `llm` and `pytest` implementation is needed; API and setup steps are unspecified.

Not counted in the score. Rated by AI (OpenAI Codex) as of Nov 15, 2026. Treat it as a guide and read the original before installing.

Disagree? Request a re-evaluation

Tell us which point is wrong and why. No login needed. We will check it against the SKILL.md, redo the evaluation, and show the result here.

This form is protected by reCAPTCHA, and the Google Privacy Policy and Terms of Service apply.

Files included (1)

  • SKILL.md5.8 KB

SKILL.md (original)

Review the instructions the agent will receive before you install.

Reviews

No reviews yet. Tell others what you thought.

Comments

No login needed. Share how it worked for you, ask a question, or note where it did not run.

No comments yet. Be the first to write one.

Comments are published right away. Do not include personal information such as email addresses or phone numbers. Up to 2 URLs.

0 / 1,000 characters

This form is protected by reCAPTCHA, and the Google Privacy Policy and Terms of Service apply.

More skills from this repository

0codekit

FreeSummary

Connect an agent to 0codekit workspace data through the Membrane CLI. The supplied guide lists many data types across spreadsheets, databases, documents, tasks, calendars, email, and other services, while noting that the exact request may need clarification.

  • Working with spreadsheet rows and columns
  • Accessing workspace tasks or calendar events
  • Handling records across listed data sources
LeoYeAI/agent-master-skills2.2KUpdated Nov 1, 2026

0xarchive

FreeSummary

Query historical and current crypto market data for Hyperliquid, Lighter.xyz, and HIP-3, including prices, candles, order books, trades, funding rates, open interest, and liquidations. Check data freshness and coverage to understand what records are available.

  • Looking up historical prices or candles for a supported exchange
  • Reviewing order book or trade history
  • Checking funding rates and open interest over time
LeoYeAI/agent-master-skills2.2KUpdated Nov 1, 2026

12-factor-apps

FreeSummary

Review a codebase against the 12-Factor App principles for SaaS architecture. Get a factor-by-factor Strong, Partial, or Weak assessment, evidence and gaps, plus prioritized recommendations for improving portability, deployment, and scaling.

  • Auditing a SaaS application's architecture
  • Reviewing a cloud-native application
  • Planning a cloud migration
LeoYeAI/agent-master-skills2.2KUpdated Nov 1, 2026

Plan personal or team projects in repeating 13-day cycles inspired by the Maya calendar. Each of 13 daily tones suggests a focus, from setting a goal and taking action to sharing results, cleaning up, and reflecting.

  • Planning a focused 13-day personal project
  • Giving a team sprint a daily focus
  • Checking the suggested focus for today
LeoYeAI/agent-master-skills2.2KUpdated Nov 1, 2026

Estimate intracellular reaction rates from steady-state carbon-13 tracing measurements. Validate carbon atom mappings and measurement inputs, simulate labeling, fit constrained fluxes, and diagnose which pathway rates the data can meaningfully constrain.

  • Estimate metabolic fluxes from carbon-13 labeling
  • Validate atom mappings and measurement inputs
  • Simulate labeling for proposed reaction rates
LeoYeAI/agent-master-skills2.2KUpdated Nov 1, 2026

Learn how to source products through 1688.com with guided steps for choosing products, checking suppliers, requesting samples, negotiating prices, and comparing shipping options. The skill also highlights common listing and payment risks and explains when an order may call for help from a sourcing expert.

  • Checking a 1688 supplier before placing an order
  • Planning product selection for a target market and budget
  • Drafting a supplier message to negotiate price or samples
LeoYeAI/agent-master-skills2.2KUpdated Nov 1, 2026

See all skills by LeoYeAI

Report a problem with this skill