PULSE-Gemini

Google Gemini adapter for PULSE Protocol — talk to Gemini with semantic messages.

Same interface as pulse-openai and pulse-anthropic — swap provider in one line. Zero Gemini boilerplate. Uses the new google-genai SDK (recommended by Google).

Quick Start

pip install pulse-gemini

from pulse import PulseMessage
from pulse_gemini import GeminiAdapter

adapter = GeminiAdapter(api_key="AIza...")

# Ask a question
msg = PulseMessage(
    action="ACT.QUERY.DATA",
    parameters={"query": "What is quantum computing?"}
)
response = adapter.send(msg)
print(response.content["parameters"]["result"])

Switch Providers in One Line

# from pulse_gemini import GeminiAdapter as Adapter
# from pulse_anthropic import AnthropicAdapter as Adapter
from pulse_openai import OpenAIAdapter as Adapter

adapter = Adapter(api_key="...")

Your code stays exactly the same. Only the import changes.

Supported Actions

PULSE Action	What It Does	Default Model
ACT.QUERY.DATA	Ask a question	gemini-2.0-flash
ACT.CREATE.TEXT	Generate text	gemini-2.0-flash
ACT.ANALYZE.SENTIMENT	Analyze sentiment	gemini-2.0-flash
ACT.ANALYZE.PATTERN	Find patterns	gemini-2.0-flash
ACT.TRANSFORM.TRANSLATE	Translate text	gemini-2.0-flash
ACT.TRANSFORM.SUMMARIZE	Summarize text	gemini-2.0-flash

Examples

Analyze sentiment

msg = PulseMessage(
    action="ACT.ANALYZE.SENTIMENT",
    parameters={"text": "This is the best day ever!"}
)
response = adapter.send(msg)

Translate text

msg = PulseMessage(
    action="ACT.TRANSFORM.TRANSLATE",
    parameters={"text": "Hello, world!", "target_language": "German"}
)
response = adapter.send(msg)

Summarize a long text

msg = PulseMessage(
    action="ACT.TRANSFORM.SUMMARIZE",
    parameters={
        "text": "Very long article text here...",
        "max_tokens": 200,
    }
)
response = adapter.send(msg)

Custom system prompt

msg = PulseMessage(
    action="ACT.QUERY.DATA",
    parameters={
        "query": "Explain recursion",
        "system_prompt": "You are a CS professor. Use simple analogies.",
    }
)
response = adapter.send(msg)

Use a different model

msg = PulseMessage(
    action="ACT.CREATE.TEXT",
    parameters={
        "instructions": "Write a poem about AI",
        "model": "gemini-2.0-pro",
        "temperature": 0.9,
    }
)
response = adapter.send(msg)

Parameters

Parameter	Description	Default
model	Override the default model	gemini-2.0-flash
temperature	Creativity (0.0 - 2.0)	0.7
max_tokens	Max response length	1000
system_prompt	Custom system instruction	per-action
target_language	For translation action	required

Why Gemini?

• Free tier — 15 requests/minute, no credit card needed
• Fast — Gemini 2.0 Flash is one of the fastest models
• Multimodal — text, images, video, audio (PULSE text actions work out of the box)
• Google ecosystem — integrates with Vertex AI for enterprise

Testing

pytest tests/ -q  # All tests mocked, no API key needed

PULSE Ecosystem

Package	Provider	Install
pulse-protocol	Core	pip install pulse-protocol
pulse-openai	OpenAI	pip install pulse-openai
pulse-anthropic	Anthropic	pip install pulse-anthropic
pulse-gemini	Google	pip install pulse-gemini
pulse-binance	Binance	pip install pulse-binance
pulse-bybit	Bybit	pip install pulse-bybit
pulse-kraken	Kraken	pip install pulse-kraken
pulse-okx	OKX	pip install pulse-okx
pulse-gateway	Gateway	pip install pulse-gateway

License

Apache 2.0 — open source, free forever.