メニュー
Top-level router for the build-paths skill set. Asks the user one question (which path?), then dispatches to beginner/, intermediate/, or advanced/. Use when the user wants to scaffold an Oracle-AI-DB project but hasn't picked a difficulty yet.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
Soccer analytics agent toolbelt. Gives Claude Code direct access to the Oracle-backed match data, ML predictions, and three-tier memory. Use when answering questions about football matches, building on the soccer agent, or exploring the World Cup dataset.
Bootstrap the soccer analytics agent workshop. Starts the Oracle AI Database Free container, applies schema, loads the FIFA dataset, optionally trains models, populates LangChain OracleVS hybrid retrieval plus semantic memory, applies LangGraph OracleDB observability, and verifies OCI GenAI access. Use when starting the workshop or resetting a stale environment.
Senior UI/UX Engineer. Architect digital interfaces overriding default LLM biases. Enforces metric-based rules, strict component architecture, CSS hardware acceleration, and balanced design engineering.
Scaffold an agent system where Oracle AI DB is the *only* state store, composed from the build-paths/skills/ building-block library. Stack — langchain-oracledb + oracle-database-mcp-server + in-DB ONNX embeddings + OCI GenAI Grok 4 + Open WebUI. Three projects — production-feeling NL2SQL+RAG hybrid analyst, self-improving research agent, conversational schema designer. For users who want a real DB-as-only-store agent demo.
Scaffold a small RAG chatbot on Oracle 26ai Free + langchain-oracledb + OCI Generative AI Grok 4 + sentence-transformers MiniLM-L6-v2 (Python-side embeddings, same model intermediate/advanced register inside Oracle) + Open WebUI. Three flavors that share one skeleton — PDF / Markdown / Web. For users new to Oracle who want a polished demo running in an afternoon.
Scaffold a Grok-4 tool-calling agent over an Oracle schema using langchain-oracledb + oracle-database-mcp-server + in-DB ONNX embeddings (registered MiniLM model, no external embedding API) + Open WebUI. For users who've built RAG before and want to rebuild it on the production-feeling Oracle stack.
SOC 職業分類に基づく
| name | build-paths |
| description | Top-level router for the build-paths skill set. Asks the user one question (which path?), then dispatches to beginner/, intermediate/, or advanced/. Use when the user wants to scaffold an Oracle-AI-DB project but hasn't picked a difficulty yet. |
| inputs | [{"target_dir":"optional; passed through to the path-specific skill"},{"topic":"optional; passed through"}] |
You are the entry point. Your only job is to pick which path's SKILL.md to hand off to. Do not scaffold anything yourself.
build-paths/README.md — what each path is, in plain English.build-paths/skills/README.md — the building-block skills the higher tiers compose.build-paths/PLAN.md — only if the user asks "what's the architecture?"You do not need to load the per-path references at this stage. Each path skill loads its own.
Print:
build-paths — pick a difficulty:
1. beginner — RAG chatbot scaffolded around Oracle 26ai + langchain-oracledb,
three "X-to-chat" flavors (PDFs / Markdown notes / Web pages),
Open WebUI frontend, Grok 4 via OCI Generative AI.
(~1 afternoon · 3 ideas to pick from)
2. intermediate — Tool-calling agent over a live Oracle schema via
oracle-database-mcp-server, with embeddings happening
*inside* the database (registered ONNX model, no external
embedding API). Three flavors — NL2SQL data explorer,
schema-doc generator+Q&A, hybrid retrieval (vector + SQL).
Open WebUI + Grok 4.
(~1-2 days · 3 ideas to pick from)
3. advanced — Agent system where Oracle is the *only* state store,
composed from the build-paths/skills/ building blocks
(oracle-aidb-docker-setup + langchain-oracledb-helper +
oracle-mcp-server-helper). Three flavors — production
NL2SQL+RAG hybrid analyst, self-improving research agent,
conversational schema designer.
(~3-5 days · 3 ideas to pick from)
Which one?
Wait for a number. If the user describes their experience instead of picking a number ("I've built RAG before"), suggest the matching path and ask them to confirm.
| Answer | Hand off to |
|---|---|
| 1 / "beginner" / "easy" / "first time with Oracle" | build-paths/beginner/SKILL.md |
| 2 / "intermediate" / "RAG" / "MCP" / "tool-calling" | build-paths/intermediate/SKILL.md |
| 3 / "advanced" / "agent" / "DB as only store" | build-paths/advanced/SKILL.md |
Pass through target_dir and topic if the user supplied them.
After handoff, the path skill takes over completely. Don't second-guess it.
The new (post-restructure) skill set requires:
docker --version works.python --version.OCI_GENAI_API_KEY (a sk-... bearer token, generated in the OCI GenAI service console). All three tiers require this; auth is bearer-token over OpenAI-compatible HTTP, NOT SigV1, so the user does NOT need ~/.oci/config or a compartment OCID. Default endpoint is https://inference.generativeai.us-phoenix-1.oci.oraclecloud.com. If the user has no API key, ask them to generate one and stop.OCI_GENAI_API_KEY must NEVER be committed — .env is in .gitignore. Confirm before writing any project files that reference it.These prerequisites apply to every tier. Don't re-ask them inside each path.
PLAN.md for the rationale and the v2 backlog." Stop.OCI_GENAI_API_KEY, a sk-... value from the OCI GenAI service console). Stop. Point them at the OCI GenAI console to generate one (no full OCI tenancy needed), or at archive/ for Ollama-flavored older ideas.README.md once; pick.