AGENT SKILL

pipecat-friday-agent

Build a low-latency, Iron Man-inspired tactical voice assistant (F.R.I.D.A.Y.) using Pipecat, Gemini, and OpenAI.

License 许可
MIT
最近更新
2026-11-10
Stars
★ 1
Forks
0

01 前言 Front Matter

name
pipecat-friday-agent
risk
safe
tags
[ "pipecat", "voice", "gemini", "openai", "python" ]
tools
[ "pipecat" ]
source
community
category
voice-agents
date_added
2026-03-10
description
Build a low-latency, Iron Man-inspired tactical voice assistant (F.R.I.D.A.Y.) using Pipecat, Gemini, and OpenAI.

02 资源文件 1

03 SKILL.md 正文

Pipecat Friday Agent

Overview

This skill provides a blueprint for building F.R.I.D.A.Y. (Replacement Integrated Digital Assistant Youth), a local voice assistant inspired by the tactical AI from the Iron Man films. It uses the Pipecat framework to orchestrate a low-latency pipeline:

  • STT: OpenAI Whisper (whisper-1) or gpt-4o-transcribe
  • LLM: Google Gemini 2.5 Flash (via a compatibility shim)
  • TTS: OpenAI TTS (nova voice)
  • Transport: Local Audio (Hardware Mic/Speakers)

When to Use This Skill

  • Use when you want to build a real-time, conversational voice agent.
  • Use when working with the Pipecat framework for pipeline-based AI.
  • Use when you need to integrate multiple providers (Google and OpenAI) into a single voice loop.
  • Use when building Iron Man-themed or tactical-themed voice applications.

How It Works

Step 1: Install Dependencies

You will need the Pipecat framework and its service providers installed:

pip install pipecat-ai[openai,google,silero] python-dotenv

Step 2: Configure Environment

Create a .env file with your API keys:

OPENAI_API_KEY=your_openai_key
GOOGLE_API_KEY=your_google_key

Step 3: Run the Agent

Execute the provided Python script to start the interface:

python scripts/friday_agent.py

Core Concepts

Pipeline Architecture

The agent follows a linear pipeline: Mic -> VAD -> STT -> LLM -> TTS -> Speaker. This allows for granular control over each stage, unlike end-to-end speech-to-speech models.

Google Compatibility Shim

Since Google's Gemini API has a different message format than OpenAI's standard (which Pipecat aggregators expect), the script includes a GoogleSafeContext and GoogleSafeMessage class to bridge the gap.

Best Practices

  • ✅ Use Silero VAD: It is robust for local hardware and prevents background noise from triggering the LLM.
  • ✅ Concise Prompts: Tactical agents should give short, data-dense responses to minimize latency.
  • ✅ Sample Rate Match: OpenAI TTS outputs at 24kHz; ensure your audio_out_sample_rate matches to avoid high-pitched or slowed audio.
  • ❌ No Polite Fillers: Avoid "Hello, how can I help you today?" Instead, use "Systems nominal. Ready for commands."

Troubleshooting

  • Problem: Audio is choppy or delayed.
    • Solution: Check your OUTPUT_DEVICE index. Run a script like test_audio_output.py to find the correct hardware index for your OS.
  • Problem: "Validation error" for message format.
    • Solution: Ensure the GoogleSafeContext shim is correctly translating OpenAI-style dicts to Gemini-style schema.

Related Skills

  • @voice-agents - General principles of voice AI.
  • @agent-tool-builder - Add tools (Search, Lights, etc.) to your Friday agent.
  • @llm-architect - Optimizing the LLM layer.

Limitations

  • Use this skill only when the task clearly matches the scope described above.
  • Do not treat the output as a substitute for environment-specific validation, testing, or expert review.
  • Stop and ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.

04 同仓库更多 Skills

YYC-Cube
YYC-Cube/YYC3-Knowledge-Base

vercel-sandbox

Run agent-browser + Chrome inside Vercel Sandbox microVMs for browser automation from any Vercel-deployed app. Use when the user needs browser automation in a Vercel app (Next.js, SvelteKit, Nuxt, Remix, Astro, etc.), wants to run headless Chrome without binary size limits, needs persistent browser sessions across commands, or wants ephemeral isolated browser environments. Triggers include "Vercel Sandbox browser", "microVM Chrome", "agent-browser in sandbox", "browser automation on Vercel", or any task requiring Chrome in a Vercel Sandbox.

★ 1
YYC-Cube
YYC-Cube/YYC3-Knowledge-Base

slack

Interact with Slack workspaces using browser automation. Use when the user needs to check unread channels, navigate Slack, send messages, extract data, find information, search conversations, or automate any Slack task. Triggers include "check my Slack", "what channels have unreads", "send a message to", "search Slack for", "extract from Slack", "find who said", or any task requiring programmatic Slack interaction.

★ 1
YYC-Cube
YYC-Cube/YYC3-Knowledge-Base

electron

Automate Electron desktop apps (VS Code, Slack, Discord, Figma, Notion, Spotify, etc.) using agent-browser via Chrome DevTools Protocol. Use when the user needs to interact with an Electron app, automate a desktop app, connect to a running app, control a native app, or test an Electron application. Triggers include "automate Slack app", "control VS Code", "interact with Discord app", "test this Electron app", "connect to desktop app", or any task requiring automation of a native Electron application.

★ 1
YYC-Cube
YYC-Cube/YYC3-Knowledge-Base

dogfood

Systematically explore and test a web application to find bugs, UX issues, and other problems. Use when asked to "dogfood", "QA", "exploratory test", "find issues", "bug hunt", "test this app/site/platform", or review the quality of a web application. Produces a structured report with full reproduction evidence -- step-by-step screenshots, repro videos, and detailed repro steps for every issue -- so findings can be handed directly to the responsible teams.

★ 1
YYC-Cube
YYC-Cube/YYC3-Knowledge-Base

agent-browser

Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction.

★ 1
YYC-Cube
YYC-Cube/YYC3-Knowledge-Base

shadcn

Manages shadcn components and projects — adding, searching, fixing, debugging, styling, and composing UI, including chat interfaces. Provides project context, component docs, and usage examples. Applies when working with shadcn/ui, component registries, presets, --preset codes, or any project with a components.json file. Also triggers for "shadcn init", "create an app with --preset", "switch to --preset", and for building a design system or applying a DESIGN.md or brand spec to shadcn/ui.

★ 1