agentsclimarketplace

Build with openai

Skill hiteshbandhu/skills-i-use/skills/ai-engineer-talks/build-with-openai

Runs OpenAI API workflows for agents, function calling, voice/Realtime, fine-tuning (SFT/DPO/RFT), sandboxes, Codex coding agents, and multimodal products. Use when building with OpenAI Agents SDK, Realtime API, tool loops, enterprise agent rollout, microVM sandboxes, or GPT-4o vision/voice.From its SKILL.md

Install
npx -y skills add hiteshbandhu/skills-i-use --skill build-with-openai

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
  • runs commandsInstructs the agent to run 4 commands, including `cp -r skills/build-with-openai ~/.claude/skills/` and 3 more.

SKILL.md

2.5 KB, 603 tokens by cl100k_base, as published. Nobody here has run it

Build with OpenAI

Action playbook from twelve OpenAI @ AI Engineer talks. Do not summarize talks — pick a workflow theme and execute it.

Supporting files:

Optional deliverables: ./skill-outputs/build-with-openai/


Step 0 — Pick workflow theme

What is the user trying to do?
├─ Tool loops, schemas, parallel calls, RAG-as-tools     → function-calling
├─ SFT / DPO / RFT, graders, shrink prompts             → fine-tuning
├─ Realtime, speech-to-speech, WebRTC vs WebSocket      → voice
├─ Isolate agent code exec, microVMs, RL rewards        → sandboxes
├─ Codex harness, Agents SDK, PR/terminal agents        → coding-agents
├─ Vision, image gen, video frames, GPT-4o modalities   → multimodal
└─ Enterprise pilot, guardrails, co-create UX           → agents-enterprise

Open the matching section in workflows.md.


Install

cp -r skills/build-with-openai ~/.claude/skills/
cp -r skills/build-with-openai ~/.cursor/skills/
cp -r skills/build-with-openai ~/.codex/skills/
cp -r skills/build-with-openai ~/.agents/skills/

Source: ingest-into-skills playlists/openai-ai-engineer/.


Cross-cutting rules

RuleSource
Agents = model + instructions + tools + runtime loop[src-007 @ 0:01]
Default 10–20 tools per turn; route if more[src-004 @ 1:16]
Eval tool execution, not just completions[src-004 @ 1:21]
Single agent in prod first; guardrails on edges[src-002 @ 12:13]
Speech-to-speech beats STT→LLM→TTS for latency[src-008 @ 0:04]

Output to user

  1. Name theme + deliverables
  2. Save artifacts under ./skill-outputs/build-with-openai/ when requested
  3. Do not auto-commit

Invocation examples

@build-with-openai design Realtime voice agent with tool delegation
RFT graders for our support bot tool routing
should we use WebRTC or WebSocket for this voice app?

What ships with it: 3 files

8.5 KB alongside SKILL.md

Gives 0 of the 12 instructions most context ai engineering skills give in 603 tokens

Counted across 1,328 of the 2,349 authors here whose files we hold, read 2026-09-06

  • Dispatch a fresh subagent for each taskin 76 of 1328, across 59 files
  • Perform spec compliance review before code quality reviewin 44 of 1328, across 34 files
  • Dispatch a final code reviewer after all tasksin 38 of 1328, across 26 files
  • Answer subagent questions before allowing implementationin 36 of 1328, across 26 files
  • Use the least powerful model capable of the taskin 33 of 1328, across 26 files
  • Create a TodoWrite list for all tasksin 32 of 1328, across 22 files
  • Perform a task review after each implementationin 31 of 1328, across 24 files
  • Extract all tasks and context from the planin 29 of 1328, across 20 files
  • Provide full task text to subagentsin 28 of 1328, across 20 files
  • Use git worktrees for isolated workspacesin 25 of 1328, across 20 files
  • Specify the model explicitly when dispatching a subagentin 23 of 1328, across 18 files
  • Execute all tasks from the plan without stoppingin 21 of 1328, across 16 files

Said here and by no other author read

  • Pick a workflow theme from the provided list
  • Open the matching section in workflows.md
  • Use speech-to-speech for voice latency
  • Implement guardrails on agent edges
  • Eval tool execution instead of just completions
  • Route requests if using over twenty tools

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 610,677. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.