القائمة
Create and run agent evals to verify AI output quality. Use when the user asks to test agent outputs, create eval cases, or benchmark skill performance.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
Full end-to-end project planning agent that takes a rough idea all the way through research, planning, organization, and validation before any code is written. Use this skill whenever a user has a new project idea and wants structured, phase-by-phase help turning it into a complete, research-backed, approved plan — even if they don't explicitly say "project planning". This is the right skill any time someone wants to think before they build. Always trigger this skill for phrases like: "I have a project idea", "help me plan my project", "let's plan this before we build", "what should I figure out before starting", "turn my idea into a project plan", "help me go from idea to plan", "project planning", "create a project brief", "I want to build a SaaS / app / tool / website — where do I start?", "I have an idea but don't know where to begin", "help me think through this idea", "should I build this?", "help me scope this project", "I need a PRD and roadmap before we code anything", "let's research this idea first
Designs and documents REST and GraphQL APIs, OpenAPI specifications, service boundaries, and integration contracts. Use this skill when a user asks to design an API, define endpoints, create a REST or GraphQL schema, write an OpenAPI or Swagger spec, establish a service contract, map integration points between services, or document an existing API. Common phrasings: "design an API for my app", "write an OpenAPI spec", "define REST endpoints", "create a GraphQL schema", "document this service's interface", "what endpoints do I need?", "help me plan my API surface". Do NOT use for implementing API server code, writing client SDK code, or running/testing live API requests — those require execution environments beyond this skill's scope.
Designs REST and GraphQL APIs including endpoint definitions, request/response schemas, authentication patterns, error-handling conventions, and OpenAPI documentation. Use this skill when a user asks to design an API, create API specs, define endpoints, document a service interface, plan an API surface, or establish request/response contracts. Common phrasings: "design an API for X", "create an OpenAPI spec", "what endpoints should I have?", "define my REST routes", "write a Swagger file", "document my service interface", "help me plan my API". Do NOT use when the user wants to implement handler code, test a live API, write client SDKs, or when the request is about database schema design (use database-schema-skill) or service orchestration architecture.
Reviews code for correctness, security vulnerabilities, performance issues, readability, and adherence to best practices and team conventions. Use this skill when a user asks for a code review, PR review, feedback on code quality, a security audit of code, style or pattern improvements, or a second opinion on an implementation. Common phrasings: "review this code", "check my PR", "look at this function", "is this secure?", "any issues with this?", "code feedback please", "audit this module", "does this follow best practices?". Do NOT use when the user wants code written from scratch (no existing code to review), or when the request is purely about running/executing code, deploying, or testing infrastructure — this skill reads and critiques existing code only.
Designs analytics and operational dashboards including layout, chart types, KPI cards, filters, drill-downs, and data visualizations. Use this skill when a user asks to design a dashboard, create data visualizations, build a metrics view, display charts or graphs, plan a reporting interface, or structure an analytics page. Common phrasings: "design a dashboard for X", "what charts should I use?", "help me visualize this data", "create a metrics view", "build a reporting page", "I need a KPI dashboard", "show me how to display this data visually". Do NOT use when the user wants to write the actual frontend code for a dashboard (use a code-generation skill), query the underlying data (use notebook-query-skill), or set up a BI tool like Tableau or Looker.
Cleans, normalizes, and validates datasets: fixes formatting inconsistencies, handles null/missing values, deduplicates records, standardizes field values, validates data types, and prepares data for downstream analysis or import. Use this skill when a user has messy data to fix, needs to normalize a CSV or table, wants data preparation before analysis or loading, or needs a cleaning plan for a dataset. Common phrasings: "clean this data", "fix my CSV", "normalize these fields", "deduplicate this table", "my data has inconsistencies", "prepare this for analysis", "standardize these values", "handle nulls in my dataset". Do NOT use when the user wants to analyze or visualize already-clean data (use notebook-query-skill), design a database schema (use database-schema-skill), or build an ETL pipeline (beyond cleaning rules, into full pipeline architecture).
| name | eval-runner |
| description | Create and run agent evals to verify AI output quality. Use when the user asks to test agent outputs, create eval cases, or benchmark skill performance. |
| category | evaluation |
| version | v1.0 |
| inputs | ["skill name","test prompts","expected outputs"] |
| outputs | ["Eval JSON files","Eval results","Pass/fail report"] |
Define, run, and report on evaluation cases for agent skills to confirm output quality and correctness.
Use this skill when the user asks to:
evals/{
"skill_name": "[name]",
"evals": [
{
"id": 1,
"name": "[eval-name]",
"prompt": "[test prompt]",
"expected_output": "[description of correct output]",
"files": []
}
]
}
evals/[skill-name]-evals.json