with one click
backend
backend contains 22 collected skills from eco2-team, with repository-level occupation coverage and site-owned skill detail pages.
Skills in this repository
Chat Eval Pipeline ์ฝ๋์ ๋ํ 5-expert ํผ๋๋ฐฑ ๋ฃจํ ๋ฆฌ๋ทฐ ๊ฐ์ด๋
Agent ์์คํ ํ๋ก๋น์ ๋ ๊ฐ์ด๋. Dependencies ์ฑ๊ธํค, HTTP ํด๋ผ์ด์ธํธ ํ๋ง, LLM ํด๋ผ์ด์ธํธ ์ค์ , ํ๊ฒฝ๋ณ์ ๊ด๋ฆฌ, Worker ์ด๊ธฐํ ํจํด. "dependencies", "provisioning", "http client", "connection pool", "worker setup" ํค์๋๋ก ํธ๋ฆฌ๊ฑฐ.
Google Gemini SDK (google-genai) ํ์ฉ ๊ฐ์ด๋. Gemini 3/2 ๋ชจ๋ธ, Structured Output, Function Calling, Google Search Grounding, ์ด๋ฏธ์ง ์์ฑ ๊ตฌํ ์ ์ฐธ์กฐ. "gemini", "google ai", "genai", "google search", "imagen" ํค์๋๋ก ํธ๋ฆฌ๊ฑฐ.
LangGraph 1.0+ ํ์ดํ๋ผ์ธ ๊ตฌ์ถ ๊ฐ์ด๋. Intent-Routed Workflow, Subagent ํจํด, Checkpointing, Conditional Routing ๊ตฌํ ์ ์ฐธ์กฐ. "langgraph", "pipeline", "graph", "workflow", "subagent", "checkpointer" ํค์๋๋ก ํธ๋ฆฌ๊ฑฐ.
LangSmith ํตํฉ ๋ฐ LLM Observability ๊ฐ์ด๋. ํ ํฐ ์ถ์ , ๋น์ฉ ๊ณ์ฐ, Run Tree, Tracing ๋ฐ์ฝ๋ ์ดํฐ, OTEL ์ฐ๋ ๊ตฌํ ์ ์ฐธ์กฐ. "langsmith", "tracing", "observability", "token usage", "cost tracking" ํค์๋๋ก ํธ๋ฆฌ๊ฑฐ.
OpenAI Agents SDK (2025~) ํ์ฉ ๊ฐ์ด๋. Agent + Runner + WebSearchTool + RunContextWrapper ํจํด. Structured Output, Function Tool, Handoff ๊ตฌํ ์ ์ฐธ์กฐ. "agents sdk", "openai agent", "runner", "web_search", "function_tool", "handoff" ํค์๋๋ก ํธ๋ฆฌ๊ฑฐ.
Guide for the Location API service (apps/location). Use when working on geospatial features, Kakao Map integration, recycling center search, nearby center queries, or location-based filtering. Triggers on "location", "kakao", "map", "geospatial", "nearby centers", "recycling centers", "store category", "pickup category".
Git Flow ๋ธ๋์น ์ ๋ต ๋ฐ PR ์ ์ฐจ ๊ฐ์ด๋. ๋ธ๋์น ์์ฑ, ์ปค๋ฐ, PR ์์ฑ ์ ์ฐธ์กฐ. "branch", "commit", "PR", "pull request", "merge", "git flow" ํค์๋๋ก ํธ๋ฆฌ๊ฑฐ.
Redis ๊ธฐ๋ฐ Composite Event Bus ํจํด ๊ฐ์ด๋. Event Router, SSE Gateway, Redis Streams/Pub-Sub ๊ตฌํ ์ ์ฐธ์กฐ. "event", "sse", "stream", "pubsub", "broadcast", "realtime" ํค์๋๋ก ํธ๋ฆฌ๊ฑฐ.
OpenAI/Gemini Function Calling์ ํ์ฉํ ์์ด์ ํธ ๊ตฌํ ๊ฐ์ด๋. LLM์ด ์ธ๋ถ API๋ฅผ Tool๋ก ํธ์ถํ๋ ํจํด. "function calling", "tool", "agent", "openai tools", "gemini tools" ํค์๋๋ก ํธ๋ฆฌ๊ฑฐ.
Ecoยฒ Chat ์๋น์ค E2E ํ ์คํธ ๋ฐ ๋๋ฒ๊น ๊ฐ์ด๋. API ์๋ํฌ์ธํธ, SSE ์ด๋ฒคํธ ์คํธ๋ฆผ, Redis Streams ํ์ดํ๋ผ์ธ, Intent ๋ถ๋ฅ ์์คํ . Use when: (1) Chat API E2E ํ ์คํธ ์, (2) SSE ์ด๋ฒคํธ ์์ ๋ฌธ์ ํด๊ฒฐ ์, (3) Intent ๋ถ๋ฅ/๋ผ์ฐํ ํ์ธ ์, (4) Redis Streams ์ด๋ฒคํธ ํ์ดํ๋ผ์ธ ๋๋ฒ๊น ์.
Chat ์๋น์ค์ ์์ํ ๋ ์ด์ด ๊ฐ์ด๋. Checkpointer(๋ฉํฐํด ์ปจํ ์คํธ), Message Persistence(PostgreSQL), Redis Streams Consumer ์ํคํ ์ฒ ๋ฐ ํธ๋ฌ๋ธ์ํ . Use when: (1) chat.messages ํ ์ด๋ธ์ด ๋น์ด์์ ๋, (2) checkpointer ๊ด๋ จ ์๋ฌ ๋ฐ์ ์, (3) ๋ฉํฐํด ๋ํ ์ปจํ ์คํธ๊ฐ ์ ์ง๋์ง ์์ ๋, (4) Redis Streams consumer group ์ค์ ์.
Celery Worker + RabbitMQ Topology CR ํจํด ๊ฐ์ด๋. ํ ์์ฑ, Exchange/Binding ์ ์, DLQ ์ค์ ์ ์ฐธ์กฐ. "celery", "rabbitmq", "queue", "topology", "dlq", "exchange", "binding" ํค์๋๋ก ํธ๋ฆฌ๊ฑฐ.
Guide for implementing Clean Architecture in Python/FastAPI projects. Use when creating new modules, refactoring existing code to Clean Architecture, designing layer structures, or implementing Port/Adapter patterns. Triggers on "clean architecture", "hexagonal", "ports and adapters", "layer structure", "DIP", "dependency inversion".
Systematic code review and evaluation tool for Python/FastAPI projects. Use when reviewing PRs, evaluating code quality, checking architecture compliance, or providing feedback on implementations. Triggers on "review", "code review", "evaluate", "check code", "PR review", "feedback".
gRPC ์๋น์ค ํตํฉ ๊ฐ์ด๋. Port/Adapter ํจํด, Proto ์ ์, Async Client ๊ตฌํ ์ ์ฐธ์กฐ. "grpc", "proto", "subagent", "character", "location", "rpc" ํค์๋๋ก ํธ๋ฆฌ๊ฑฐ.
Kubernetes ์ด์/๋๋ฒ๊น ๊ฐ์ด๋. Pod ํธ๋ฌ๋ธ์ํ , ๋ก๊ทธ ๋ถ์, ๋คํธ์ํฌ ์ง๋จ, ํด๋ฌ์คํฐ ๊ตฌ์กฐ, GitOps/ArgoCD, Operator/CRD/CR, Helm ๋ฐฐํฌ ์ ์ฐธ์กฐ. "k8s", "kubernetes", "pod", "debug", "logs", "kubectl", "argocd", "helm", "operator", "crd" ํค์๋๋ก ํธ๋ฆฌ๊ฑฐ.
PostgreSQL ์คํค๋ง ์ค๊ณ ๊ฐ์ด๋. DDL ์์ฑ, ํ ์ด๋ธ ์ค๊ณ, ์ธ๋ฑ์ค ์ ๋ต, ๋ง์ด๊ทธ๋ ์ด์ ์ ์ฐธ์กฐ. "schema", "ddl", "table", "migration", "index", "constraint" ํค์๋๋ก ํธ๋ฆฌ๊ฑฐ.
ํ๋กฌํํธ ์์ง๋์ด๋ง ๊ฐ์ด๋. ํ๋กฌํํธ ํ ํ๋ฆฟ ์ค๊ณ, ํ๊ฐ, ์ต์ ํ ์ ์ฐธ์กฐ. "prompt", "template", "system prompt", "few-shot", "chain-of-thought" ํค์๋๋ก ํธ๋ฆฌ๊ฑฐ.
RAG (Retrieval-Augmented Generation) ํ์ดํ๋ผ์ธ ๊ตฌํ ๊ฐ์ด๋. Contextual Retrieval, Evidence Tracking, Phase 1-4 ํ๊ฐ ์ ์ฐธ์กฐ. "rag", "retrieval", "retriever", "evidence", "citation", "nugget" ํค์๋๋ก ํธ๋ฆฌ๊ฑฐ.
Redis ํจํด ๊ฐ์ด๋. Cache-Aside, Rate Limiting, Pub/Sub, Streams ๊ตฌํ ์ ์ฐธ์กฐ. "redis", "cache", "rate limit", "pubsub", "streams" ํค์๋๋ก ํธ๋ฆฌ๊ฑฐ.
Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.